Network Simulation and Testing
description
Transcript of Network Simulation and Testing
![Page 2: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/2.jpg)
2
Traffic Papers• V. Paxson, and S. Floyd, Wide-Area Traffic: The Failure of Poisson
Modeling. IEEE/ACM Transactions on Networking, Vol. 3 No. 3, pp. 226-244, June 1995
• W. E. Leland, M. S. Taqqu, W. Willinger, and D. V. Wilson, On the Self-Similar Nature of Ethernet Traffic. IEEE/ACM Transactions on Networking, Vol. 2, No. 1, pp. 1-15, Feb. 1995
• M. E. Crovella and A. Bestavros, Self-Similarity in World Wide Web Traffic: Evidence and Possible Causes. IEEE/ACM Transactions on Networking, Vol 5, No. 6, pp. 835-846, December 1997
• Anja Feldmann; Anna C. Gilbert; Polly Huang; Walter Willinger, Dynamics of IP traffic: A study of the role of variability and the impact of control. In the Proceeding of SIGCOMM '99, Cambridge, Massachusetts, September 1999
![Page 3: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/3.jpg)
3
Paper Selection
Interesting Boring Easy Difficult
Failure of PoissonSS in EthernetSS in WWWIP Dynamics
![Page 4: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/4.jpg)
4
Identifying Internet Traffic
Failure of PoissonSelf-similar Traffic
Practical Model
![Page 5: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/5.jpg)
5
The Problem
• What is the traffic workload like?
• Call/packet arrival rate as a process• What kind of process is it?
• Very old problem and a lot of work
![Page 6: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/6.jpg)
6
Because
• Traces are available • Researchers care about
– The validness of their assumption– The network traffic being independent Poisson
• Operation people care a lot about – The amount of buffer/bandwidth to provision for their
networks– The profit comes from satisfying customers with
minimum infrastructure cost
![Page 7: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/7.jpg)
7
Telephone Network
• Assumptions– Poisson call arrivals– Exponential call duration
• Wonderful Property– Poisson mixing with Poisson is still Poisson– Average rate well-characterize a call
• The whole queueing theory
![Page 8: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/8.jpg)
8
Data Network?
• Wide-Area Traffic: The Failure of Poisson Modeling
• V. Paxson, and S. Floyd• IEEE/ACM Transactions on Networking,
Vol. 3 No. 3, pp. 226-244, June 1995
![Page 9: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/9.jpg)
9
A Study of the Wide-Area Traffic
• Two units of examination– Connections vs. packets
• A sizeable number of traces– ~4M connections, ~26M packets– Different location and different time
• Inter-arrival processes– TCP connections– Telnet packets– FTPDATA connections
• Going self-similar
![Page 10: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/10.jpg)
10
Unit of Observation
• Telephone network– Circuit-switched– The unit is circuit, i.e., a call– People picking up the phone and talk for a while
• Data network– Packet-switched– The unit is packet– Another unit is connection, comparable to call– People starting up an FTP connection and send data for
a while
![Page 11: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/11.jpg)
11
Packet Connection
• Hosts send/receive packets over a channel at the transport layer– Reliable: TCP– Non-reliable: UDP
• Packets from various channels multiplex at the the network layer– IP Routers switched on the packets
![Page 12: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/12.jpg)
12
Inter-Arrival Process: A Little Exercise
Beginning SYNACK&SYN
ACK&Segment 1
End
FIN
ACK&FINACK
Beginning
?
?
![Page 13: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/13.jpg)
13
TCP Connection Arrival Poisson?
Depends
![Page 14: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/14.jpg)
14
Application Dependent• TELNET
– Users typing ‘telnet cc.ee.ntu.edu.tw’• FTP
– User typing ‘ftp cc.ee.ntu.edu.tw’• FTPDATA burst
– User typing ‘mget net-simtest-*.ppt’• FTPDATA
– Each individual TCP transfer• NNTP & SMTP
– Machine initiated and/or timer-driven
![Page 15: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/15.jpg)
15
Independent and Poisson?
Y/N
TELNET
FTP
FTPDATA
FTPDATA burst
SMTP
NNTP
![Page 16: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/16.jpg)
16
Quick Summary
• TELNET and FTP– Independent and Poisson– Both the 1-hour and 10-min scales
• FTPDATA bursts and SMTP– At the 10-min interval– Not ‘terribly far’ from Poisson– SMTP inter-arrival is not independent
• FTPDATA, NNTP– Clearly not Poisson
![Page 17: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/17.jpg)
17
Before One Can Explain
• Human-initiated process– Independent and Poisson
• Non-human-initiated process– Well, who knows
![Page 18: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/18.jpg)
18
Explanations I
• TELNET and FTP– User initiated– Users typing ‘telnet cc.ee.ntu.edu.tw’– User typing ‘ftp cc.ee.ntu.edu.tw’
• FTPDATA bursts– User typing ‘mget net-simtest-*.ppt’– Actually, taking the closely-spaced connections… (<= 4
sec)• FTPDATA
– TCP connections
![Page 19: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/19.jpg)
19
Explanations II
• NNTP– Flooding to propagate network news– Arrival of news trigger another– Periodical and implementation/configuration
dependent• SMTP
– Mailing list– Timer effects from the DNS queries
![Page 20: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/20.jpg)
20
TELNET Packets Poisson?
No, heavy-tailed!
![Page 21: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/21.jpg)
21
Show in 4 Ways
• Distribution of packet inter-arrival time– Exponential processes ramp up significantly slower
• Packet arrival pattern in seconds and 10 seconds– Exponential processes are smoother at the 10sec scale
• Variance-time plot– Change of variance to time scale– Var of exponential processes decays quickly
• Packet arrival rate process in seconds– By the sole visual effect– Exponential processes are less spiky
![Page 22: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/22.jpg)
22
Full TELNET model?
Poisson connection arrivalHeavy-tailed packet arrival within a connection
![Page 23: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/23.jpg)
23
FTPDATA
• Connection arrival is not Poisson– Clustered in bursts
• Burst sizes in bytes is quite heavy-tailed– A 0.5% of bursts contribute to 50% of the
traffic volume
![Page 24: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/24.jpg)
24
OK. We know it’s not Poisson. But what?
![Page 25: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/25.jpg)
25
Going Self-Similar
• Well, since other evidences suggest so • And it’s the next good thing
• Go straight into producing self-similar traffic
![Page 26: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/26.jpg)
26
Producing Self-Similar Traffic
• ON/OFF sources– Fix ON period rate– ON/OFF period length heavy-tailed
• M/G/– Customer arrival being Poisson– Service time being heavy-tailed with infinite variance
• Authors’ own model– Pseudo-self-similar– Not long-range dependent though
![Page 27: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/27.jpg)
27
Performance Implication
• Low-priority traffic starvation– Shall the high-priority traffic being long-range
dependent (bursty)• Admission control based on recent traffic
failing– ‘Congestions haven’t happened for a long
while’ does not mean it won’t happen now
![Page 28: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/28.jpg)
28
The Real Message
Poisson is no longer sufficient!
![Page 29: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/29.jpg)
29
Identifying Internet Traffic
Failure of PoissonSelf-similar Traffic
Practical Model
![Page 30: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/30.jpg)
30
Self-Similar What?
• On the Self-Similar Nature of Ethernet Traffic
• Will E. Leland; Murad S. Taqqu; Walter Willinger; Daniel V. Wilson
• IEEE/ACM Transactions on Networking, Vol. 2, No. 1, pp. 1-15, Feb. 1995
![Page 31: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/31.jpg)
31
This One Easier
• Self-similarity in World Wide Web Traffic: Evidence and Possible Causes
• Mark E. Crovella; Azer Bestavros• IEEE/ACM Transactions on Networking,
Vol 5, No. 6, pp. 835-846, December 1997
![Page 32: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/32.jpg)
32
Self-Similar Process
Serpgask Triangles
![Page 33: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/33.jpg)
33
Definition
• X: a stationary time series• X(m): the m-aggregates
– Summing the time series over non-overlapping blocks of m
• X is H-self-similar if– X (m) has the same distribution for all positive m
![Page 34: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/34.jpg)
34
Same Distribution?
• Same autocorrelation function– r(k) = E[(Xt - )(Xt+k - )]/2
• r(k) ~ k-
– k – 0 < < 1
![Page 35: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/35.jpg)
35
Significance of k-
• Long-range dependence– Just another way of characterizing the same thing
• Power-law decay– Slower than exponential decay– Therefore traffic does not smooth up
< 1– r(k) does not converge– Sum of r(k) infinite, I.e., variance infinite
![Page 36: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/36.jpg)
36
Just FYI
• The Hurst parameter: 1- /2
![Page 37: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/37.jpg)
37
Tests for Self-Similarity• Variance-time plot
– A line with slope - > -1• R/S plot
– Rescaled range grows as the number points included– A line with slope H an the log-log scale
• Periodogram– Power spectrum to frequency– A line with slope - 1 at the log-log scale
• Whittle estimator– Confidence to a form– FGN or Fractional ARIMA
![Page 38: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/38.jpg)
38
Pareto Review
• Exponential– f(x) = ce-cx
• Heavy-tailed– F(x) ~ x-c, 0 < c < 2– Hyperbolic
• Pareto– f(x) = ckc x-c-1
– F(x) = 1- (k/x)c
– A line at the log-log scale of F(x) plot
![Page 39: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/39.jpg)
39
In Addition to the Theory
• A HUGE volume of Ethernet traces• Show consistency of being self-similar in
all sorts of tests• Implication to traffic engineering
• A bombshell!
![Page 40: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/40.jpg)
40
Why Self-Similar?
• Theory suggests– Fix rate ON/OFF process– Heavy-tailed length
• Looking into the length– The ON time: transmission time– The OFF time: silent time
![Page 41: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/41.jpg)
41
Physical Cause
• Heavy-tailed transmission time– Heavy-tailed file sizes– Magic of the nature – E.g., book size in library
![Page 42: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/42.jpg)
42
Identifying Internet Traffic
Failure of PoissonSelf-similar TrafficPractical Model
![Page 43: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/43.jpg)
43
So, enough Math. Just tell me what to do!
It depends!
![Page 44: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/44.jpg)
44
Cutting to the Chase
• The structural model– user level: Poisson arrival and heavy-tailed
duration– network level: TCP closed-loop feedback
control and ack clocking– Variability: delay and congestion
• Let simulators track the complex behavior
![Page 45: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/45.jpg)
45
Why not FGN?
• IP Traffic Dynamics: The Role of Variability and Control
• Anja Feldmann; Anna C. Gilbert; Polly Huang; Walter Willinger
• In the Proceeding of SIGCOMM '99, Cambridge, Massachusetts, September 1999
![Page 46: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/46.jpg)
46
Remember Wavelet Analysis?
• FFT – Frequency decomposition– fj, Fourier coefficient– Amount of the signal in frequency j
• WT: wavelet transform– Frequency (scale) and time decomposition– dj,k, wavelet coefficient– Amount of the signal in frequency j, time k
![Page 47: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/47.jpg)
47
Self-similarity
• Energy function– Ej = Σ(dj,k)2/Nj
– Weighted average of the signal strength at scale j
• Self-similar process– Ej = 2j(2H-1) C <- the magic!!– log2 Ej = (2H-1) j + log2C– linear relationship between log2 Ej and j
![Page 48: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/48.jpg)
48
‘Shape’ of Self-Similarity
Self-similar
?? RTT
![Page 49: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/49.jpg)
49
Wavelet Example
0-1
1
00 00 00 00 11 11 11 11s1
s2
s3
s4
d1
d2
d3
d4
0 0 0 0 2 2 2 2 0 0 0 0 0 0 0 00 0 4 4 0 0 0 0
0 8 0 08 8
![Page 50: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/50.jpg)
50
Adding Periodicity
• packets arrive periodically, 1 pkt/23 msec• coefficients cancel out at scale 4
10 00 00 00 10 00 00 00s1
s2
s3
s4
d1
d2
d3
d4
1 0 0 0 1 0 0 0 1 0 0 0 1 0 0 01 0 1 0 1 0 1 0
1 1 1 12 0
![Page 51: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/51.jpg)
51
Visualization
J=4
Adding Periodicity
![Page 52: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/52.jpg)
52
’Shape' of self-similarity
Self-similar
Yes RTT!
![Page 53: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/53.jpg)
53
Large Scale
• Heavy-tailed connection duration
![Page 54: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/54.jpg)
54
Medium Scale
• TCP close-loop control
RTT
![Page 55: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/55.jpg)
55
TCP Flow Control
source sink
RTT RTT RTTTime
![Page 56: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/56.jpg)
56
Variability
• Delay and congestion (bandwidth & load)Simulation Measurement
![Page 57: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/57.jpg)
57
Internet Traffic is Weird!
• Different properties at different time scales– Large scales: self-similarity– Medium scale: periodicity– Small scale: ??? (possibly multifractal)
![Page 58: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/58.jpg)
58
New Queuing Theory?
• For chaotic Internet traffic• Only pen and paper
![Page 59: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/59.jpg)
59
NO!
• Probably not in the near future• Confirmed by the experts
![Page 60: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/60.jpg)
60
A Few Reasons
• Not exactly self-similar (FGN - big no no)• ’Shape' of self-similarity changes with the
network conditions• Don't know what self-similar processes add
up to (mathematically intractable)• Don’t know what those strange small-scale
behavior is exactly
![Page 61: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/61.jpg)
61
Therefore
• The structural model– User level: Poisson arrival and heavy-tailed
duration– Network level: TCP closed-loop feedback
control and ack clocking– Variability: delay and congestion
• Let simulators track the complex behavior
![Page 62: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/62.jpg)
62
Questions?
![Page 63: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/63.jpg)
63
On the Review Forms
• Novelty– New idea
• Clarity– The problem
• Reality (practicality)– Evaluation
• Importance, significance, relevance– How much impact?– Would things change?
![Page 64: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/64.jpg)
64
OK for Beginners
• Clarity– Easiest– Judging the writing
• Evaluation– Easy– Judging the experiments and technical content
![Page 65: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/65.jpg)
65
Challenging for the Advanced
• Novelty– Hard– Need to follow/read enough papers in the area
• Importance– Hardest– Need to have breadth and know enough
development in the area
![Page 66: Network Simulation and Testing](https://reader036.fdocuments.in/reader036/viewer/2022062501/5681602d550346895dcf419a/html5/thumbnails/66.jpg)
66
Show your FreeBSD installation!