Transport Layer By Ossi Mokryn, Based also on slides from: the Computer Networking: A Top Down...
-
Upload
amari-reams -
Category
Documents
-
view
214 -
download
1
Transcript of Transport Layer By Ossi Mokryn, Based also on slides from: the Computer Networking: A Top Down...
Transport Layer
By Ossi Mokryn Based also on slides from the Computer Networking
A Top Down Approach Featuring the Internet by Kurose and Ross
Transport Layer
Connectionless and connection oriented communication
Sockets programming UDP TCP
Reliable communication Flow control Congestion control Timers
2Transport Layer
Transport services and protocols provide logical
communication between app processes running on different hosts
transport protocols run in end systems send side breaks app
messages into segments passes to network layer
rcv side reassembles segments into messages passes to app layer
more than one transport protocol available to apps Internet TCP and UDP
application
transportnetworkdata linkphysical
application
transportnetworkdata linkphysical
logical end-end transport
Transport Layer 3
Internet transport-layer protocols reliable in-order
delivery (TCP) congestion control flow control connection setup
unreliable unordered delivery UDP no-frills extension of
ldquobest-effortrdquo IP services not available
delay guarantees bandwidth guarantees
application
transportnetworkdata linkphysical network
data linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
application
transportnetworkdata linkphysical
logical end-end transport
Transport Layer 4
Transport vs network layer
network layer logical communication between hosts
transport layer logical communication between processes relies on enhances network layer services
Household analogy12 kids sending letters
to 12 kids processes = kids app messages =
letters in envelopes hosts = houses transport protocol =
Ann and Bill network-layer
protocol = postal service
Transport Layer 5
Multiplexingdemultiplexing
application
transport
network
link
physical
P1 application
transport
network
link
physical
application
transport
network
link
physical
P2P3 P4P1
host 1 host 2 host 3
= process= socket
delivering received segmentsto correct socket
Demultiplexing at rcv hostgathering data from multiplesockets enveloping data with header (later used for demultiplexing)
Multiplexing at send host
Transport Layer 6
How demultiplexing works host receives IP datagrams
each datagram has source IP address destination IP address
each datagram carries 1 transport-layer segment
each segment has source destination port number
host uses IP addresses amp port numbers to direct segment to appropriate socket
source port dest port
32 bits
applicationdata
(message)
other header fields
TCPUDP segment format
Transport Layer 7
Connectionless demultiplexing Create sockets with port
numbers UDP socket identified by
two-tuple(dest IP address dest port number)
When host receives UDP segment checks destination port
number in segment directs UDP segment to
socket with that port number
IP datagrams with different source IP addresses andor source port numbers directed to same socket
Transport Layer 8
Connectionless demux (cont)
DatagramSocket serverSocket = new DatagramSocket(6428)
ClientIPB
P2
client IP A
P1P1P3
serverIP C
SP 6428
DP 9157
SP 9157
DP 6428
SP 6428
DP 5775
SP 5775
DP 6428
SP provides ldquoreturn addressrdquo
Transport Layer 9
UDP User Datagram Protocol [RFC 768]
Simplest Internet transport protocol
Each app Output produces exactly one UDP segment
ldquobest effortrdquo service UDP segments may be lost delivered out of order to
app connectionless
no handshaking between UDP sender receiver
each UDP segment handled independently of others
Why is there a UDP no connection
establishment (which can add delay)
simple no connection state at sender receiver
small segment header no congestion control
UDP can blast away as fast as desired
Transport Layer 10
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Transport Layer
Connectionless and connection oriented communication
Sockets programming UDP TCP
Reliable communication Flow control Congestion control Timers
2Transport Layer
Transport services and protocols provide logical
communication between app processes running on different hosts
transport protocols run in end systems send side breaks app
messages into segments passes to network layer
rcv side reassembles segments into messages passes to app layer
more than one transport protocol available to apps Internet TCP and UDP
application
transportnetworkdata linkphysical
application
transportnetworkdata linkphysical
logical end-end transport
Transport Layer 3
Internet transport-layer protocols reliable in-order
delivery (TCP) congestion control flow control connection setup
unreliable unordered delivery UDP no-frills extension of
ldquobest-effortrdquo IP services not available
delay guarantees bandwidth guarantees
application
transportnetworkdata linkphysical network
data linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
application
transportnetworkdata linkphysical
logical end-end transport
Transport Layer 4
Transport vs network layer
network layer logical communication between hosts
transport layer logical communication between processes relies on enhances network layer services
Household analogy12 kids sending letters
to 12 kids processes = kids app messages =
letters in envelopes hosts = houses transport protocol =
Ann and Bill network-layer
protocol = postal service
Transport Layer 5
Multiplexingdemultiplexing
application
transport
network
link
physical
P1 application
transport
network
link
physical
application
transport
network
link
physical
P2P3 P4P1
host 1 host 2 host 3
= process= socket
delivering received segmentsto correct socket
Demultiplexing at rcv hostgathering data from multiplesockets enveloping data with header (later used for demultiplexing)
Multiplexing at send host
Transport Layer 6
How demultiplexing works host receives IP datagrams
each datagram has source IP address destination IP address
each datagram carries 1 transport-layer segment
each segment has source destination port number
host uses IP addresses amp port numbers to direct segment to appropriate socket
source port dest port
32 bits
applicationdata
(message)
other header fields
TCPUDP segment format
Transport Layer 7
Connectionless demultiplexing Create sockets with port
numbers UDP socket identified by
two-tuple(dest IP address dest port number)
When host receives UDP segment checks destination port
number in segment directs UDP segment to
socket with that port number
IP datagrams with different source IP addresses andor source port numbers directed to same socket
Transport Layer 8
Connectionless demux (cont)
DatagramSocket serverSocket = new DatagramSocket(6428)
ClientIPB
P2
client IP A
P1P1P3
serverIP C
SP 6428
DP 9157
SP 9157
DP 6428
SP 6428
DP 5775
SP 5775
DP 6428
SP provides ldquoreturn addressrdquo
Transport Layer 9
UDP User Datagram Protocol [RFC 768]
Simplest Internet transport protocol
Each app Output produces exactly one UDP segment
ldquobest effortrdquo service UDP segments may be lost delivered out of order to
app connectionless
no handshaking between UDP sender receiver
each UDP segment handled independently of others
Why is there a UDP no connection
establishment (which can add delay)
simple no connection state at sender receiver
small segment header no congestion control
UDP can blast away as fast as desired
Transport Layer 10
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Transport services and protocols provide logical
communication between app processes running on different hosts
transport protocols run in end systems send side breaks app
messages into segments passes to network layer
rcv side reassembles segments into messages passes to app layer
more than one transport protocol available to apps Internet TCP and UDP
application
transportnetworkdata linkphysical
application
transportnetworkdata linkphysical
logical end-end transport
Transport Layer 3
Internet transport-layer protocols reliable in-order
delivery (TCP) congestion control flow control connection setup
unreliable unordered delivery UDP no-frills extension of
ldquobest-effortrdquo IP services not available
delay guarantees bandwidth guarantees
application
transportnetworkdata linkphysical network
data linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
application
transportnetworkdata linkphysical
logical end-end transport
Transport Layer 4
Transport vs network layer
network layer logical communication between hosts
transport layer logical communication between processes relies on enhances network layer services
Household analogy12 kids sending letters
to 12 kids processes = kids app messages =
letters in envelopes hosts = houses transport protocol =
Ann and Bill network-layer
protocol = postal service
Transport Layer 5
Multiplexingdemultiplexing
application
transport
network
link
physical
P1 application
transport
network
link
physical
application
transport
network
link
physical
P2P3 P4P1
host 1 host 2 host 3
= process= socket
delivering received segmentsto correct socket
Demultiplexing at rcv hostgathering data from multiplesockets enveloping data with header (later used for demultiplexing)
Multiplexing at send host
Transport Layer 6
How demultiplexing works host receives IP datagrams
each datagram has source IP address destination IP address
each datagram carries 1 transport-layer segment
each segment has source destination port number
host uses IP addresses amp port numbers to direct segment to appropriate socket
source port dest port
32 bits
applicationdata
(message)
other header fields
TCPUDP segment format
Transport Layer 7
Connectionless demultiplexing Create sockets with port
numbers UDP socket identified by
two-tuple(dest IP address dest port number)
When host receives UDP segment checks destination port
number in segment directs UDP segment to
socket with that port number
IP datagrams with different source IP addresses andor source port numbers directed to same socket
Transport Layer 8
Connectionless demux (cont)
DatagramSocket serverSocket = new DatagramSocket(6428)
ClientIPB
P2
client IP A
P1P1P3
serverIP C
SP 6428
DP 9157
SP 9157
DP 6428
SP 6428
DP 5775
SP 5775
DP 6428
SP provides ldquoreturn addressrdquo
Transport Layer 9
UDP User Datagram Protocol [RFC 768]
Simplest Internet transport protocol
Each app Output produces exactly one UDP segment
ldquobest effortrdquo service UDP segments may be lost delivered out of order to
app connectionless
no handshaking between UDP sender receiver
each UDP segment handled independently of others
Why is there a UDP no connection
establishment (which can add delay)
simple no connection state at sender receiver
small segment header no congestion control
UDP can blast away as fast as desired
Transport Layer 10
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Internet transport-layer protocols reliable in-order
delivery (TCP) congestion control flow control connection setup
unreliable unordered delivery UDP no-frills extension of
ldquobest-effortrdquo IP services not available
delay guarantees bandwidth guarantees
application
transportnetworkdata linkphysical network
data linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
networkdata linkphysical
application
transportnetworkdata linkphysical
logical end-end transport
Transport Layer 4
Transport vs network layer
network layer logical communication between hosts
transport layer logical communication between processes relies on enhances network layer services
Household analogy12 kids sending letters
to 12 kids processes = kids app messages =
letters in envelopes hosts = houses transport protocol =
Ann and Bill network-layer
protocol = postal service
Transport Layer 5
Multiplexingdemultiplexing
application
transport
network
link
physical
P1 application
transport
network
link
physical
application
transport
network
link
physical
P2P3 P4P1
host 1 host 2 host 3
= process= socket
delivering received segmentsto correct socket
Demultiplexing at rcv hostgathering data from multiplesockets enveloping data with header (later used for demultiplexing)
Multiplexing at send host
Transport Layer 6
How demultiplexing works host receives IP datagrams
each datagram has source IP address destination IP address
each datagram carries 1 transport-layer segment
each segment has source destination port number
host uses IP addresses amp port numbers to direct segment to appropriate socket
source port dest port
32 bits
applicationdata
(message)
other header fields
TCPUDP segment format
Transport Layer 7
Connectionless demultiplexing Create sockets with port
numbers UDP socket identified by
two-tuple(dest IP address dest port number)
When host receives UDP segment checks destination port
number in segment directs UDP segment to
socket with that port number
IP datagrams with different source IP addresses andor source port numbers directed to same socket
Transport Layer 8
Connectionless demux (cont)
DatagramSocket serverSocket = new DatagramSocket(6428)
ClientIPB
P2
client IP A
P1P1P3
serverIP C
SP 6428
DP 9157
SP 9157
DP 6428
SP 6428
DP 5775
SP 5775
DP 6428
SP provides ldquoreturn addressrdquo
Transport Layer 9
UDP User Datagram Protocol [RFC 768]
Simplest Internet transport protocol
Each app Output produces exactly one UDP segment
ldquobest effortrdquo service UDP segments may be lost delivered out of order to
app connectionless
no handshaking between UDP sender receiver
each UDP segment handled independently of others
Why is there a UDP no connection
establishment (which can add delay)
simple no connection state at sender receiver
small segment header no congestion control
UDP can blast away as fast as desired
Transport Layer 10
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Transport vs network layer
network layer logical communication between hosts
transport layer logical communication between processes relies on enhances network layer services
Household analogy12 kids sending letters
to 12 kids processes = kids app messages =
letters in envelopes hosts = houses transport protocol =
Ann and Bill network-layer
protocol = postal service
Transport Layer 5
Multiplexingdemultiplexing
application
transport
network
link
physical
P1 application
transport
network
link
physical
application
transport
network
link
physical
P2P3 P4P1
host 1 host 2 host 3
= process= socket
delivering received segmentsto correct socket
Demultiplexing at rcv hostgathering data from multiplesockets enveloping data with header (later used for demultiplexing)
Multiplexing at send host
Transport Layer 6
How demultiplexing works host receives IP datagrams
each datagram has source IP address destination IP address
each datagram carries 1 transport-layer segment
each segment has source destination port number
host uses IP addresses amp port numbers to direct segment to appropriate socket
source port dest port
32 bits
applicationdata
(message)
other header fields
TCPUDP segment format
Transport Layer 7
Connectionless demultiplexing Create sockets with port
numbers UDP socket identified by
two-tuple(dest IP address dest port number)
When host receives UDP segment checks destination port
number in segment directs UDP segment to
socket with that port number
IP datagrams with different source IP addresses andor source port numbers directed to same socket
Transport Layer 8
Connectionless demux (cont)
DatagramSocket serverSocket = new DatagramSocket(6428)
ClientIPB
P2
client IP A
P1P1P3
serverIP C
SP 6428
DP 9157
SP 9157
DP 6428
SP 6428
DP 5775
SP 5775
DP 6428
SP provides ldquoreturn addressrdquo
Transport Layer 9
UDP User Datagram Protocol [RFC 768]
Simplest Internet transport protocol
Each app Output produces exactly one UDP segment
ldquobest effortrdquo service UDP segments may be lost delivered out of order to
app connectionless
no handshaking between UDP sender receiver
each UDP segment handled independently of others
Why is there a UDP no connection
establishment (which can add delay)
simple no connection state at sender receiver
small segment header no congestion control
UDP can blast away as fast as desired
Transport Layer 10
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Multiplexingdemultiplexing
application
transport
network
link
physical
P1 application
transport
network
link
physical
application
transport
network
link
physical
P2P3 P4P1
host 1 host 2 host 3
= process= socket
delivering received segmentsto correct socket
Demultiplexing at rcv hostgathering data from multiplesockets enveloping data with header (later used for demultiplexing)
Multiplexing at send host
Transport Layer 6
How demultiplexing works host receives IP datagrams
each datagram has source IP address destination IP address
each datagram carries 1 transport-layer segment
each segment has source destination port number
host uses IP addresses amp port numbers to direct segment to appropriate socket
source port dest port
32 bits
applicationdata
(message)
other header fields
TCPUDP segment format
Transport Layer 7
Connectionless demultiplexing Create sockets with port
numbers UDP socket identified by
two-tuple(dest IP address dest port number)
When host receives UDP segment checks destination port
number in segment directs UDP segment to
socket with that port number
IP datagrams with different source IP addresses andor source port numbers directed to same socket
Transport Layer 8
Connectionless demux (cont)
DatagramSocket serverSocket = new DatagramSocket(6428)
ClientIPB
P2
client IP A
P1P1P3
serverIP C
SP 6428
DP 9157
SP 9157
DP 6428
SP 6428
DP 5775
SP 5775
DP 6428
SP provides ldquoreturn addressrdquo
Transport Layer 9
UDP User Datagram Protocol [RFC 768]
Simplest Internet transport protocol
Each app Output produces exactly one UDP segment
ldquobest effortrdquo service UDP segments may be lost delivered out of order to
app connectionless
no handshaking between UDP sender receiver
each UDP segment handled independently of others
Why is there a UDP no connection
establishment (which can add delay)
simple no connection state at sender receiver
small segment header no congestion control
UDP can blast away as fast as desired
Transport Layer 10
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
How demultiplexing works host receives IP datagrams
each datagram has source IP address destination IP address
each datagram carries 1 transport-layer segment
each segment has source destination port number
host uses IP addresses amp port numbers to direct segment to appropriate socket
source port dest port
32 bits
applicationdata
(message)
other header fields
TCPUDP segment format
Transport Layer 7
Connectionless demultiplexing Create sockets with port
numbers UDP socket identified by
two-tuple(dest IP address dest port number)
When host receives UDP segment checks destination port
number in segment directs UDP segment to
socket with that port number
IP datagrams with different source IP addresses andor source port numbers directed to same socket
Transport Layer 8
Connectionless demux (cont)
DatagramSocket serverSocket = new DatagramSocket(6428)
ClientIPB
P2
client IP A
P1P1P3
serverIP C
SP 6428
DP 9157
SP 9157
DP 6428
SP 6428
DP 5775
SP 5775
DP 6428
SP provides ldquoreturn addressrdquo
Transport Layer 9
UDP User Datagram Protocol [RFC 768]
Simplest Internet transport protocol
Each app Output produces exactly one UDP segment
ldquobest effortrdquo service UDP segments may be lost delivered out of order to
app connectionless
no handshaking between UDP sender receiver
each UDP segment handled independently of others
Why is there a UDP no connection
establishment (which can add delay)
simple no connection state at sender receiver
small segment header no congestion control
UDP can blast away as fast as desired
Transport Layer 10
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Connectionless demultiplexing Create sockets with port
numbers UDP socket identified by
two-tuple(dest IP address dest port number)
When host receives UDP segment checks destination port
number in segment directs UDP segment to
socket with that port number
IP datagrams with different source IP addresses andor source port numbers directed to same socket
Transport Layer 8
Connectionless demux (cont)
DatagramSocket serverSocket = new DatagramSocket(6428)
ClientIPB
P2
client IP A
P1P1P3
serverIP C
SP 6428
DP 9157
SP 9157
DP 6428
SP 6428
DP 5775
SP 5775
DP 6428
SP provides ldquoreturn addressrdquo
Transport Layer 9
UDP User Datagram Protocol [RFC 768]
Simplest Internet transport protocol
Each app Output produces exactly one UDP segment
ldquobest effortrdquo service UDP segments may be lost delivered out of order to
app connectionless
no handshaking between UDP sender receiver
each UDP segment handled independently of others
Why is there a UDP no connection
establishment (which can add delay)
simple no connection state at sender receiver
small segment header no congestion control
UDP can blast away as fast as desired
Transport Layer 10
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Connectionless demux (cont)
DatagramSocket serverSocket = new DatagramSocket(6428)
ClientIPB
P2
client IP A
P1P1P3
serverIP C
SP 6428
DP 9157
SP 9157
DP 6428
SP 6428
DP 5775
SP 5775
DP 6428
SP provides ldquoreturn addressrdquo
Transport Layer 9
UDP User Datagram Protocol [RFC 768]
Simplest Internet transport protocol
Each app Output produces exactly one UDP segment
ldquobest effortrdquo service UDP segments may be lost delivered out of order to
app connectionless
no handshaking between UDP sender receiver
each UDP segment handled independently of others
Why is there a UDP no connection
establishment (which can add delay)
simple no connection state at sender receiver
small segment header no congestion control
UDP can blast away as fast as desired
Transport Layer 10
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
UDP User Datagram Protocol [RFC 768]
Simplest Internet transport protocol
Each app Output produces exactly one UDP segment
ldquobest effortrdquo service UDP segments may be lost delivered out of order to
app connectionless
no handshaking between UDP sender receiver
each UDP segment handled independently of others
Why is there a UDP no connection
establishment (which can add delay)
simple no connection state at sender receiver
small segment header no congestion control
UDP can blast away as fast as desired
Transport Layer 10
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
UDP more
often used for streaming multimedia apps loss tolerant rate sensitive
other UDP uses DNS SNMP
reliable transfer over UDP add reliability at application layer application-specific
error recovery
source port dest port
32 bits
Applicationdata
(message)
UDP segment format
length checksumLength in
bytes of UDPsegmentincluding
header and dataMinimum value is 8
Bytes
Transport Layer 11
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
UDP checksum
Sender treat segment contents
as sequence of 16-bit integers
checksum addition (1rsquos complement sum) of segment contents
sender puts checksum value into UDP checksum field
Receiver compute checksum of
received segment check if computed
checksum equals checksum field value NO - error detected YES - no error detected
But maybe errors nonetheless More later hellip
Goal detect ldquoerrorsrdquo (eg flipped bits) in transmitted segment
Transport Layer 12
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Internet Checksum Example Note
When adding numbers a carryout from the most significant bit needs to be added to the result
Example add two 16-bit integers
1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 01 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1
1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 0 01 0 1 0 0 0 1 0 0 0 1 0 0 0 0 1 1
wraparound
sumchecksum
Transport Layer 13
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Sockets Programming over UDP ndash use socket slides now
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Transmission Control Protocol
Principles of reliable communication TCP basic notations 3 way handshake TCP flow control congestion control
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 16
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 17
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Principles of Reliable data transfer important in app transport link layers top-10 list of important networking topics
characteristics of unreliable channel will determine complexity of reliable data transfer protocol (rdt)
Transport Layer 18
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Reliable Data Transfer Stream stream jargon
Application Layer 19
A stream is a sequence of characters that flow into or out of a process
An input stream is attached to some input source for the process eg keyboard or socket
An output stream is attached to an output source eg monitor or socket
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Reliable Communication
20Transport Layer
state1
state2
event causing state transitionactions taken on state transition
eventactions
Terminology of a State Machine
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Reliable communication
First Model sender sends receiver receives
Is this enough When will it work When will it not work
Transport Layer 21
Sender sends one packet then waits for receiver response
stop and wait
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Reliable Communication
22Transport Layer
State1
state2
Send Data
Data available
Stop and Wait ndash Sender side
Received AckIn State 1
Sender can send data
In State 2 Sender can receive acknowledge packets
Discussion bull Why to send back an ack msgbull What happens if data is available at state
2
Wait
for data Wait
for ack
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
channel with bit errors and losses
underlying channel may flip bits in packet checksum to detect bit errors
underlying channel can also lose packets the question how to recover from errors
acknowledgements (ACKs) receiver explicitly tells sender that pkt received OK
timeout sender retransmits pkt if doesnrsquot receive ack within timeout
new mechanisms in error detection receiver feedback control msg (ACK) rcvr-gtsender sender control timer to understand if to send again
Transport Layer 23
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Reliable Communication
24Transport Layer
State1
state2a
Send DataData available
Stop and Wait with errorslosses - sender
Ack receivedIn State 1
Receiver waits for data
Wait for data
Wait for ack packet or time out
state2b
timeout
Send D
ata
Discussion bull What does the sender need to do for
the retransmission
Process ack
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
This version has a fatal flaw
What happens if ACK corruptedlost sender doesnrsquot know what happened at receiver canrsquot just retransmit possible duplicate
Handling duplicates sender retransmits
current pkt if ACK garbled or didnrsquot arrive
sender adds sequence number to each pkt
receiver discards (doesnrsquot deliver up) duplicate pkt
receiver must specify seq of pkt being ACKed
Transport Layer 25
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
discussionSender seq added to pkt two seq rsquos (01)
will suffice Why must check if
received ACK corrupted
twice as many states state must
ldquorememberrdquo whether ldquocurrentrdquo pkt has 0 or 1 seq
Receiver must check if received
packet is duplicate state indicates whether 0
or 1 is expected pkt seq receiver sends ACK for
last pkt received OK receiver must explicitly
include seq of pkt being ACKed
note receiver can not know if its last ACK received OK at sender
Transport Layer 26
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Stop amp wait in action
Transport Layer 27
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Stop amp wait in action
Transport Layer 28
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Performance of stop amp wait
Stop amp wait works but performance stinks ex 1 Gbps link 15 ms prop delay 8000 bit
packet
U sender utilization ndash fraction of time sender busy sending
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
1KB pkt every 30 msec -gt 33kBsec thruput over 1 Gbps link
network protocol limits use of physical resources
dsmicrosecon8bps10
bits80009
R
Ldtrans
Transport Layer 29
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
stop-and-wait operation
first packet bit transmitted t = 0
sender receiver
RTT
last packet bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
U sender
= 008
30008 = 000027
microseconds
L R
RTT + L R =
Transport Layer 30
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Pipelined protocols
Pipelining sender allows multiple ldquoin-flightrdquo yet-to-be-acknowledged pkts range of sequence numbers must be increased buffering at sender andor receiver
Two generic forms of pipelined protocols go-Back-N selective repeat
Transport Layer 31
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Pipelining increased utilization
first packet bit transmitted t = 0
sender receiver
RTT
last bit transmitted t = L R
first packet bit arriveslast packet bit arrives send ACK
ACK arrives send next packet t = RTT + L R
last bit of 2nd packet arrives send ACKlast bit of 3rd packet arrives send ACK
U sender
= 024
30008 = 00008
microseconds
3 L R
RTT + L R =
Increase utilizationby a factor of 3
Transport Layer 32
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Pipelining Protocol
Transport Layer 33
Go-back-N big picture Sender can have up to N unacked packets in
pipeline Rcvr only sends cumulative acks
Doesnrsquot ack packet if therersquos a gap Sender has timer for oldest unacked packet
If timer expires retransmit all unacked packets
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Transport Layer 3-34
Go-Back-NSender k-bit seq in pkt header ldquowindowrdquo of up to N consecutive unackrsquoed pkts allowed
ACK(n) ACKs all pkts up to including seq n - ldquocumulative ACKrdquo may receive duplicate ACKs (see receiver)
timer for each in-flight pkt timeout(n) retransmit pkt n and all higher seq pkts in
window
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Go Back N
Transport Layer 35
ACK-only always send ACK for correctly-received pkt with highest in-order seq may generate duplicate ACKs need only remember expectedseqnum
out-of-order pkt discard (donrsquot buffer) -gt no receiver buffering Re-ACK pkt with highest in-order seq
Receiver
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Transport Layer 3-36
GBN inaction
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Transport Control Protocol
Enhanced GBN protocol
Segment structure reliable data transfer and data transfer
issues flow control connection management TCP congestion control
Transport Layer 37
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Overview RFCs 793 1122 1323 2018 2581
full duplex data bi-directional data flow
in same connection MSS maximum
segment size connection-oriented
handshaking (exchange of control msgs) initrsquos sender receiver state before data exchange
flow controlled sender will not
overwhelm receiver
point-to-point one sender one
receiver reliable in-order byte
steam no ldquomessage
boundariesrdquo pipelined
TCP congestion and flow control set window size
send amp receive buffers
socketdoor
T C Psend buffer
T C Preceive buffer
socketdoor
segm ent
applicationwrites data
applicationreads data
Transport Layer 38
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP segment structure
source port dest port
32 bits
applicationdata
(variable length)
sequence number
acknowledgement numberReceive window
Urg data pnterchecksum
FSRPAUheadlen
notused
Options (variable length)
URG urgent data (generally not used)
ACK ACK valid
PSH push data now(generally not used)
RST SYN FINconnection estab(setup teardown
commands)
bytes rcvr willingto accept
countingby bytes of data(not segments)
Internetchecksum
(as in UDP)
Transport Layer 39
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP reliable data transfer
TCP creates reliable service on top of IPrsquos unreliable service
Pipelined segments Cumulative acks TCP uses single retransmission timer
The sequence number for a segment is the first byte-stream of the first byte in the segment
Transport Layer 40
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP sender eventsdata rcvd from app Create segment with
seq start timer if not
already running (think of timer as for oldest unacked segment)
expiration interval TimeOutInterval
timeout retransmit segment
that caused timeout restart timer Ack rcvd If acknowledges
previously unacked segments update what is known
to be acked start timer if there are
outstanding segments
Transport Layer 41
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP seq rsquos and ACKsSeq rsquos
byte stream ldquonumberrdquo of first byte in segmentrsquos data
ACKs seq of next
byte expected from other side
cumulative ACKQ how receiver
handles out-of-order segments A TCP spec
doesnrsquot say - up to implementor
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Usertypes
lsquoCrsquo
host receivesACK
host ACKsreceipt of
lsquoCrsquo
time
Transport Layer 42
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP retransmission scenarios
Host A
Seq=100 20 bytes data
ACK=100
time
premature timeout
Host B
Seq=92 8 bytes data
ACK=120
Seq=92 8 bytes data
Seq=
92
tim
eout
ACK=120
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
lost ACK scenario
Host B
X
Seq=92 8 bytes data
ACK=100
time
Seq=
92
tim
eout
SendBase= 100
SendBase= 120
SendBase= 120
Sendbase= 100
Transport Layer 43
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP retransmission scenarios (more)
Host A
Seq=92 8 bytes data
ACK=100
loss
tim
eout
Cumulative ACK scenario
Host B
X
Seq=100 20 bytes data
ACK=120
time
SendBase= 120
Transport Layer 44
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Interactive data flow Overhead for each
packet 40 bytes (20 TCP header + 20 IP header) to a total of 160 bytes for sending and receiving lsquoCrsquo
If the receiver waits a while it can piggyback the data packet
Delayed ack Wait up to 500ms for next segment If no next segment send ACK
Should sender use delayed acks too[Stevens figure 193]
Host A Host B
Seq=42 ACK=79 data = lsquoCrsquo
Seq=79 ACK=43
Seq=43 ACK=80
Usertypes
lsquoCrsquo
host ACKsreceipt
of echoedlsquoCrsquo
host ACKsreceipt of
lsquoCrsquo
timesimple telnet scenario
Transport Layer 45
Seq=79 ACK=43 data = lsquoCrsquo
Host echoesback lsquoCrsquo
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Nagle Algorithm [RFC 896]
Quantifying overhead how much control bytes per data bytes with piggyback 2120-gt only 16 of the bits sent are data
LANs usually not congested so it might be okay
Small packets termed tinygrams over congested WAN ndash bad news
New data canrsquot be sent until outstanding data is acked
Small amounts of data are collected and sent in a single segment when ack arrives
Self clocking the faster the ack comes back the faster data is sent Slow links cause fewer segments to be sent
[Stevens 194]
Transport Layer 46
Naglersquos alg
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP ACK generation [RFC 1122 RFC 2581]
Event at Receiver
Arrival of in-order segment withexpected seq All data up toexpected seq already ACKed
Arrival of in-order segment withexpected seq One other segment has ACK pending
Arrival of out-of-order segmenthigher-than-expect seq Gap detected
Arrival of segment that partially or completely fills gap
TCP Receiver action
Delayed ACK Wait up to 500msfor next segment If no next segmentsend ACK
Immediately send single cumulative ACK ACKing both in-order segments
Immediately send duplicate ACK indicating seq of next expected byte
Immediate send ACK provided thatsegment starts at lower end of gap
Transport Layer 47
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Fast Retransmit
Time-out period often relatively long long delay before resending lost packet
Detect lost segments via duplicate ACKs Sender often sends many segments back-to-back If segment is lost there will likely be many duplicate
ACKs
If sender receives 3 ACKs for the same data it supposes that segment after ACKed data was lost fast retransmit resend
segment before timer expires
Transport Layer 48
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Host A
tim
eout
Host B
time
X
resend 2nd segment
Figure 337 Resending a segment after triple duplicate ACKTransport Layer 49
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
event ACK received with ACK field value of y if (y gt SendBase) SendBase = y if (there are currently not-yet-acknowledged segments) start timer else increment count of dup ACKs received for y if (count of dup ACKs received for y = 3) resend segment with sequence number y
Fast retransmit algorithm
a duplicate ACK for already ACKed segment
fast retransmit
Transport Layer 50
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Round Trip Time and TimeoutQ how to set TCP
timeout value longer than RTT
but RTT varies too short premature
timeout unnecessary
retransmissions too long slow
reaction to segment loss
Q how to estimate RTT SampleRTT measured time
from segment transmission until ACK receipt ignore retransmissions
SampleRTT will vary want estimated RTT ldquosmootherrdquo average several recent
measurements not just current SampleRTT
Transport Layer 51
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Round Trip Time and TimeoutEstimatedRTT = (1- )EstimatedRTT + SampleRTT
Exponential weighted moving average influence of past sample decreases exponentially
fast typical value = 0125
Transport Layer 52
[Retransmission example in Stevens 211]
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Example RTT estimationRTT gaiacsumassedu to fantasiaeurecomfr
100
150
200
250
300
350
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106
time (seconnds)
RTT
(mill
isec
onds
)
SampleRTT Estimated RTT
Transport Layer 53
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Round Trip Time and TimeoutSetting the timeout EstimtedRTT plus ldquosafety marginrdquo
large variation in EstimatedRTT -gt larger safety margin first estimate of how much SampleRTT deviates from
EstimatedRTT
TimeoutInterval = EstimatedRTT + 4DevRTT
DevRTT = (1-)DevRTT + |SampleRTT-EstimatedRTT|
(typically = 025)
Then set timeout interval
Transport Layer 54
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Flow Control
receive side of TCP connection has a receive buffer
speed-matching service matching the send rate to the receiving apprsquos drain rate app process may be
slow at reading from buffer
sender wonrsquot overflow
receiverrsquos buffer bytransmitting too
much too fast
flow control
Transport Layer 55
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Flow control how it works
(Suppose TCP receiver discards out-of-order segments)
spare room in buffer= RcvWindow= RcvBuffer-[LastByteRcvd -
LastByteRead]
Rcvr advertises spare room by including value of RcvWindow in segments
Sender limits unACKed data to RcvWindow guarantees receive
buffer doesnrsquot overflow
Discuss [Stevens 201 206]
Transport Layer 56
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Connection Management
Recall TCP sender receiver establish ldquoconnectionrdquo before exchanging data segments
initialize TCP variables seq s buffers flow control
info (eg RcvWindow) client connection
initiator Socket clientSocket = new
Socket(hostnameport
number) server contacted by
client Socket connectionSocket =
welcomeSocketaccept()
Three way handshake
Step 1 client host sends TCP SYN segment to server specifies initial seq no data
Step 2 server host receives SYN replies with SYNACK segment server allocates buffers specifies server initial
seq Step 3 client receives
SYNACK replies with ACK segment which may contain data
Transport Layer 57
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Delayed Duplicates Problem
A user asks for a connection Due to congestion the packet is caught in a traffic jam The user asks again for the connection Destination accepts 2nd connection request User sends info to dest Info gets caught in a traffic jam User sends info again Dest receives the info Connection is closed by both parties The original connection request and user info find their
way to the destination
Transport Layer 58
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Connection Management (cont)
Closing a connection
client closes socket clientSocketclose()
Step 1 client end system sends TCP FIN control segment to server
Step 2 server receives FIN replies with ACK Closes connection sends FIN
client
FIN
server
ACK
ACK
FIN
close
close
closed
tim
ed w
ait
Transport Layer 59
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Connection Management (cont)
Step 3 client receives FIN replies with ACK
Enters ldquotimed waitrdquo - will respond with ACK to received FINs
Step 4 server receives ACK Connection closed
Note with small modification can handle simultaneous FINs
client
FIN
server
ACK
ACK
FIN
closing
closing
closed
tim
ed w
ait
closed
Transport Layer 60
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Connection Management (cont)
TCP clientlifecycle
TCP serverlifecycle
Transport Layer 61
[Tanenbaum 633]
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Principles of Congestion Control
Congestion informally ldquotoo many sources sending too
much data too fast for network to handlerdquo different from flow control manifestations
lost packets (buffer overflow at routers) long delays (queueing in router buffers)
a top-10 problem
Transport Layer 62
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Causescosts of congestion scenario 1
two senders two receivers
one router infinite buffers
no retransmission
large delays when congested
maximum achievable throughput
unlimited shared output link buffers
Host Alin original data
Host B
lout
Transport Layer 63
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Causescosts of congestion scenario 2
one router finite buffers sender retransmission of lost packet
finite shared output link buffers
Host A lin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 64
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Causescosts of congestion scenario 2 always (goodput)
ldquoperfectrdquo retransmission only when loss
retransmission of delayed (not lost) packet makes
larger (than perfect case) for same
lin
lout
=
lin
lout
gtl
inl
out
ldquocostsrdquo of congestion more work (retrans) for given ldquogoodputrdquo unneeded retransmissions link carries multiple copies of
pkt
R2
R2lin
l ou
t
b
R2
R2lin
l ou
t
a
R2
R2lin
l ou
t
c
R4
R3
Transport Layer 65
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Causescosts of congestion scenario 3 four senders multihop paths timeoutretransmit
lin
Q what happens as and increase
lin
finite shared output link buffers
Host Alin original data
Host B
lout
lin original data plus retransmitted data
Transport Layer 66
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Causescosts of congestion scenario 3
Another ldquocostrdquo of congestion when packet dropped any ldquoupstream
transmission capacity used for that packet was wasted
Host A
Host B
lo
u
t
Transport Layer 67
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP congestion control additive increase multiplicative decrease
8 Kbytes
16 Kbytes
24 Kbytes
time
congestionwindow
Approach increase transmission rate (window size) probing for usable bandwidth until loss occurs additive increase increase CongWin by 1
MSS every RTT until loss detected multiplicative decrease cut CongWin in half
after loss
timecong
estio
n w
indo
w s
ize
Saw toothbehavior probing
for bandwidth
Transport Layer 68
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Congestion Control details
sender limits transmission LastByteSent-LastByteAcked CongWin Roughly
CongWin is dynamic function of perceived network congestion
How does sender perceive congestion
loss event = timeout or 3 duplicate acks
TCP sender reduces rate (CongWin) after loss event
three mechanisms AIMD slow start conservative after
timeout events
rate = CongWin
RTT Bytessec
Transport Layer 69
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Slow Start
When connection begins CongWin = 1 MSS Example MSS = 500
bytes amp RTT = 200 msec
initial rate = 20 kbps available bandwidth
may be gtgt MSSRTT desirable to quickly
ramp up to respectable rate
When connection begins increase rate exponentially fast until first loss event
Transport Layer 70
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Slow Start (more)
When connection begins increase rate exponentially until first loss event double CongWin every RTT done by incrementing CongWin for every ACK
received Summary initial rate is slow but ramps up
exponentially fast
Host A
one segment
RTT
Host B
time
two segments
four segments
Transport Layer 71
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Refinement inferring loss
After 3 dup ACKs CongWin is cut in half window then grows
linearly But after timeout event
CongWin instead set to 1 MSS
window then grows exponentially
to a threshold then grows linearly
3 dup ACKs indicates
network capable of delivering some segments
timeout indicates a ldquomore alarmingrdquo congestion scenario
Philosophy
Transport Layer 72
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Refinement
Q When should the exponential increase switch to linear
A When CongWin gets to 12 of its value before timeout
Implementation Variable Threshold At loss event Threshold
is set to 12 of CongWin just before loss event
Transport Layer 73
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Summary TCP Congestion Control
When CongWin is below Threshold sender in slow-start phase window grows exponentially
When CongWin is above Threshold sender is in congestion-avoidance phase window grows linearly
When a triple duplicate ACK occurs Threshold set to CongWin2 and CongWin set to Threshold
When timeout occurs Threshold set to CongWin2 and CongWin is set to 1 MSS
Transport Layer 74
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP sender congestion control
State Event TCP Sender Action Commentary
Slow Start (SS)
ACK receipt for previously unacked data
CongWin = CongWin + MSS If (CongWin gt Threshold) set state to ldquoCongestion Avoidancerdquo
Resulting in a doubling of CongWin every RTT
CongestionAvoidance (CA)
ACK receipt for previously unacked data
CongWin = CongWin+MSS (MSSCongWin)
Additive increase resulting in increase of CongWin by 1 MSS every RTT
SS or CA Loss event detected by triple duplicate ACK
Threshold = CongWin2 CongWin = ThresholdSet state to ldquoCongestion Avoidancerdquo
Fast recovery implementing multiplicative decrease CongWin will not drop below 1 MSS
SS or CA Timeout Threshold = CongWin2 CongWin = 1 MSSSet state to ldquoSlow Startrdquo
Enter slow start
SS or CA Duplicate ACK
Increment duplicate ACK count for segment being acked
CongWin and Threshold not changed
Transport Layer 75
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP throughput
Whatrsquos the average throughout of TCP as a function of window size and RTT Ignore slow start
Let W be the window size when loss occurs
When window is W throughput is WRTT Just after loss window drops to W2
throughput to W2RTT Average throughout 75 WRTT
Transport Layer 76
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
TCP Futures TCP over ldquolong fat pipesrdquo
Example 1500 byte segments 100ms RTT want 10 Gbps throughput
Requires window size W = 83333 in-flight segments
Throughput in terms of loss rate
L = 210-10 Wow New versions of TCP for high-speed
LRTT
MSS221
Transport Layer 77
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Fairness goal if K TCP sessions share same bottleneck link of bandwidth R each should have average rate of RK
TCP connection 1
bottleneckrouter
capacity R
TCP connection 2
TCP Fairness
Transport Layer 78
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Why is TCP fair
Two competing sessions Additive increase gives slope of 1 as throughout increases multiplicative decrease decreases throughput proportionally
R
R
equal bandwidth share
Connection 1 throughputConnect
ion 2
th
roughput
congestion avoidance additive increaseloss decrease window by factor of 2
congestion avoidance additive increaseloss decrease window by factor of 2
Transport Layer 79
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Fairness (more)
Fairness and UDP Multimedia apps
often do not use TCP do not want rate
throttled by congestion control
Instead use UDP pump audiovideo at
constant rate tolerate packet loss
Research area TCP friendly
Fairness and parallel TCP connections
nothing prevents app from opening parallel connections between 2 hosts
Web browsers do this Example link of rate R
supporting 9 connections new app asks for 1 TCP
gets rate R10 new app asks for 11 TCPs
gets R2
Transport Layer 80
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Chapter 3 Summary principles behind
transport layer services multiplexing
demultiplexing reliable data transfer flow control congestion control
instantiation and implementation in the Internet UDP TCP
After that leaving the
network ldquoedgerdquo (application transport layers)
into the network ldquocorerdquo
Transport Layer 81
Next Socket
programming over TCP
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Connection-oriented demux
TCP socket identified by 4-tuple source IP address source port number dest IP address dest port number
recv host uses all four values to direct segment to appropriate socket
Server host may support many simultaneous TCP sockets each socket identified
by its own 4-tuple Web servers have
different sockets for each connecting client non-persistent HTTP will
have different socket for each request
Transport Layer 82
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Connection-oriented demux (cont)
ClientIPB
P1
client IP A
P1P2P4
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P5 P6 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 83
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-
Connection-oriented demux Threaded Web Server
ClientIPB
P1
client IP A
P1P2
serverIP C
SP 9157
DP 80
SP 9157
DP 80
P4 P3
D-IPCS-IP A
D-IPC
S-IP B
SP 5775
DP 80
D-IPCS-IP B
Transport Layer 84
- Transport Layer
- Transport Layer (2)
- Transport services and protocols
- Internet transport-layer protocols
- Transport vs network layer
- Multiplexingdemultiplexing
- How demultiplexing works
- Connectionless demultiplexing
- Connectionless demux (cont)
- UDP User Datagram Protocol [RFC 768]
- UDP more
- UDP checksum
- Internet Checksum Example
- Sockets Programming over UDP ndash use socket slides now
- Transmission Control Protocol
- Principles of Reliable data transfer
- Principles of Reliable data transfer (2)
- Principles of Reliable data transfer (3)
- Reliable Data Transfer Stream stream jargon
- Reliable Communication
- Reliable communication
- Reliable Communication (2)
- channel with bit errors and losses
- Reliable Communication (3)
- This version has a fatal flaw
- discussion
- Stop amp wait in action
- Stop amp wait in action (2)
- Performance of stop amp wait
- stop-and-wait operation
- Pipelined protocols
- Pipelining increased utilization
- Pipelining Protocol
- Go-Back-N
- Go Back N
- GBN in action
- Transport Control Protocol
- TCP Overview RFCs 793 1122 1323 2018 2581
- TCP segment structure
- TCP reliable data transfer
- TCP sender events
- TCP seq rsquos and ACKs
- TCP retransmission scenarios
- TCP retransmission scenarios (more)
- Interactive data flow
- Nagle Algorithm [RFC 896]
- TCP ACK generation [RFC 1122 RFC 2581]
- Fast Retransmit
- Slide 49
- Fast retransmit algorithm
- TCP Round Trip Time and Timeout
- TCP Round Trip Time and Timeout (2)
- Example RTT estimation
- TCP Round Trip Time and Timeout (3)
- TCP Flow Control
- TCP Flow control how it works
- TCP Connection Management
- Delayed Duplicates Problem
- TCP Connection Management (cont)
- TCP Connection Management (cont) (2)
- TCP Connection Management (cont)
- Principles of Congestion Control
- Causescosts of congestion scenario 1
- Causescosts of congestion scenario 2
- Causescosts of congestion scenario 2 (2)
- Causescosts of congestion scenario 3
- Causescosts of congestion scenario 3 (2)
- TCP congestion control additive increase multiplicative decre
- TCP Congestion Control details
- TCP Slow Start
- TCP Slow Start (more)
- Refinement inferring loss
- Refinement
- Summary TCP Congestion Control
- TCP sender congestion control
- TCP throughput
- TCP Futures TCP over ldquolong fat pipesrdquo
- TCP Fairness
- Why is TCP fair
- Fairness (more)
- Chapter 3 Summary
- Connection-oriented demux
- Connection-oriented demux (cont)
- Connection-oriented demux Threaded Web Server
-