Fall 2006 CS 561 1

Peer-to-Peer Networks

OutlineOverview

Pastry

OpenDHT

Contributions from Peter Druschel & Sean Rhea

Fall 2006 CS 561 2

What is Peer-to-Peer About?

• Distribution• Decentralized control• Self-organization• Symmetric communication

• P2P networks do two things:– Map objects onto nodes

– Route requests to the node responsible for a given object

Fall 2006 CS 561 4

Object Distribution

objId

Consistent hashing [Karger et al. ‘97]

128 bit circular id space

nodeIds (uniform random)

objIds (uniform random)

Invariant: node with numerically closest nodeId maintains object

nodeIds

O2128 - 1

Fall 2006 CS 561 5

Object Insertion/Lookup

X

Route(X)

Msg with key X is routed to live node with nodeId closest to X

Problem:

complete routing table not feasible

O2128 - 1

Fall 2006 CS 561 6

Routing: Distributed Hash Table

Properties• log16 N steps • O(log N) state

d46a1c

Route(d46a1c)

d462ba

d4213f

d13da3

65a1fc

d467c4d471f1

Fall 2006 CS 561 7

Leaf Sets

Each node maintains IP addresses of the nodes with the L numerically closest larger and smaller nodeIds, respectively. • routing efficiency/robustness • fault detection (keep-alive)• application-specific local coordination

Fall 2006 CS 561 8

Routing Procedureif (D is within range of our leaf set)

forward to numerically closest memberelse

let l = length of shared prefix let d = value of l-th digit in D’s addressif (RouteTab[l,d] exists)

forward to RouteTab[l,d]else

forward to a known node that (a) shares at least as long a prefix(b) is numerically closer than this node

Fall 2006 CS 561 9

Routing

Integrity of overlay:• guaranteed unless L/2 simultaneous failures of

nodes with adjacent nodeIds

Number of routing hops:

• No failures: < log16 N expected, 128/b + 1 max

• During failure recovery:– O(N) worst case, average case much better

Fall 2006 CS 561 10

Node Addition

d46a1c

Route(d46a1c)

d462ba

d4213f

d13da3

65a1fc

d467c4d471f1

New node: d46a1c

Fall 2006 CS 561 11

Node Departure (Failure)

Leaf set members exchange keep-alive messages

• Leaf set repair (eager): request set from farthest live node in set

• Routing table repair (lazy): get table from peers in the same row, then higher rows

Fall 2006 CS 561 12

PAST: Cooperative, archival file storage and distribution

• Layered on top of Pastry• Strong persistence• High availability• Scalability• Reduced cost (no backup)• Efficient use of pooled resources

Fall 2006 CS 561 13

PAST API

• Insert - store replica of a file at k diverse storage nodes

• Lookup - retrieve file from a nearby live storage node that holds a copy

• Reclaim - free storage associated with a file

Files are immutable

Fall 2006 CS 561 14

PAST: File storage

fileId

Insert fileId

Fall 2006 CS 561 15

PAST: File storage

Storage Invariant: File “replicas” are stored on k nodes with nodeIdsclosest to fileId

(k is bounded by the leaf set size)

fileId

Insert fileId

k=4

Fall 2006 CS 561 16

PAST: File Retrieval

fileId file located in log16 N steps (expected)

usually locates replica nearest client C

Lookup

k replicasC

Fall 2006 CS 561 17

IP

ChordDHT

CFS(MIT)

PastryDHT

PAST(MSR/Rice)

TapestryDHT

OStore(UCB)

BambooDHT

PIER(UCB)

CANDHT

pSearch(HP)

KademliaDHT

Coral(NYU)

ChordDHT

i3(UCB)

DHT Deployment Today

KademliaDHT

Overnet(open)

connectivity

Every application deploys its own DHT(DHT as a library)

Fall 2006 CS 561 18

IP

ChordDHT

CFS(MIT)

PastryDHT

PAST(MSR/Rice)

TapestryDHT

OStore(UCB)

BambooDHT

PIER(UCB)

CANDHT

pSearch(HP)

KademliaDHT

Coral(NYU)

ChordDHT

i3(UCB)

KademliaDHT

Overnet(open)

DHT

connectivity

indirection

OpenDHT: one DHT, shared across applications(DHT as a service)

DHT Deployment Tomorrow?

Fall 2006 CS 561 19

Two Ways To Use a DHT

1. The Library Model– DHT code is linked into application binary

– Pros: flexibility, high performance

2. The Service Model– DHT accessed as a service over RPC

– Pros: easier deployment, less maintenance

Fall 2006 CS 561 20

The OpenDHT Service

• 200-300 Bamboo [USENIX’04] nodes on PlanetLab– All in one slice, all managed by us

• Clients can be arbitrary Internet hosts– Access DHT using RPC over TCP

• Interface is simple put/get:– put(key, value) — stores value under key

– get(key) — returns all the values stored under key

• Running on PlanetLab since April 2004– Building a community of users

Fall 2006 CS 561 21

OpenDHT ApplicationsApplication Uses OpenDHT for

Croquet Media Manager replica location

DOA indexing

HIP name resolution

DTN Tetherless Computing Architecture host mobility

Place Lab range queries

QStream multicast tree construction

VPN Index indexing

DHT-Augmented Gnutella Client rare object search

FreeDB storage

Instant Messaging rendezvous

CFS storage

i3 redirection

Fall 2006 CS 561 22

An Example Application: The CD Database

Compute Disc Fingerprint

Recognize Fingerprint?

Album & Track Titles

Fall 2006 CS 561 23

An Example Application: The CD Database

Type In Album andTrack Titles

Album & Track Titles

No Such Fingerprint

Fall 2006 CS 561 24

A DHT-Based FreeDB Cache

• FreeDB is a volunteer service– Has suffered outages as long as 48 hours

– Service costs born largely by volunteer mirrors

• Idea: Build a cache of FreeDB with a DHT– Add to availability of main service

– Goal: explore how easy this is to do

Fall 2006 CS 561 25

Cache IllustrationDHTDHT

New AlbumsDisc

Fingerprin

t

Disc In

fo

Disc Fingerprint

Fall 2006 CS 561 26

Is Providing DHT Service Hard?

• Is it any different than just running Bamboo?– Yes, sharing makes the problem harder

• OpenDHT is shared in two senses– Across applications need a flexible interface

– Across clients need resource allocation

Fall 2006 CS 561 27

Sharing Between Applications

• Must balance generality and ease-of-use– Many apps want only simple put/get

– Others want lookup, anycast, multicast, etc.

• OpenDHT allows only put/get– But use client-side library, ReDiR, to build others

– Supports lookup, anycast, multicast, range search

– Only constant latency increase on average

– (Different approach used by DimChord [KR04])

Fall 2006 CS 561 28

Sharing Between Clients

• Must authenticate puts/gets/removes– If two clients put with same key, who wins?

– Who can remove an existing put?

• Must protect system’s resources– Or malicious clients can deny service to others

– The remainder of this talk

Fall 2006 CS 561 29

Fair Storage Allocation

• Our solution: give each client a fair share– Will define “fairness” in a few slides

• Limits strength of malicious clients– Only as powerful as they are numerous

• Protect storage on each DHT node separately– Global fairness is hard– Key choice imbalance is a burden on DHT– Reward clients that balance their key choices

Fall 2006 CS 561 30

Two Main Challenges

1. Making sure disk is available for new puts– As load changes over time, need to adapt

– Without some free disk, our hands are tied

2. Allocating free disk fairly across clients– Adapt techniques from fair queuing

Fall 2006 CS 561 31

Making Sure Disk is Available

• Can’t store values indefinitely– Otherwise all storage will eventually fill

• Add time-to-live (TTL) to puts– put (key, value) put (key, value, ttl)

– (Different approach used by Palimpsest [RH03])

Fall 2006 CS 561 32

Making Sure Disk is Available

• TTLs prevent long-term starvation– Eventually all puts will expire

• Can still get short term starvation:

time

Client A arrivesfills entire of disk

Client B arrivesasks for space

Client A’s valuesstart expiring

B Starves

Fall 2006 CS 561 33

Making Sure Disk is Available

• Stronger condition:Be able to accept rmin bytes/sec new data at all times

Reserved for futureputs. Slope = rmin

Candidate put

TTL

size

Sum must be < max capacity

time

spac

e

max

max0now

Fall 2006 CS 561 34

Making Sure Disk is Available

• Stronger condition:Be able to accept rmin bytes/sec new data at all times

TTL

size

time

spac

e

max

max0now

TTLsize

time

spac

e

max

max0now

Fall 2006 CS 561 35

Fair Storage Allocation

Per-clientput queues

Queue full:reject put

Not full:enqueue put

Select mostunder-

represented

Wait until canaccept withoutviolating rmin

Store andsend accept

message to client

The Big Decision: Definition of “most under-represented”

Fall 2006 CS 561 36

Defining “Most Under-Represented”• Not just sharing disk, but disk over time

– 1-byte put for 100s same as 100-byte put for 1s– So units are bytes seconds, call them commitments

• Equalize total commitments granted?– No: leads to starvation– A fills disk, B starts putting, A starves up to max TTL

time

Client A arrivesfills entire of disk

Client B arrivesasks for space

B catches up with A

Now A Starves!

Fall 2006 CS 561 37

Defining “Most Under-Represented”• Instead, equalize rate of commitments granted

– Service granted to one client depends only on others putting “at same time”

time

Client A arrivesfills entire of disk

Client B arrivesasks for space

B catches up with A

A & B shareavailable rate

Fall 2006 CS 561 38

Defining “Most Under-Represented”• Instead, equalize rate of commitments granted

– Service granted to one client depends only on others putting “at same time”

• Mechanism inspired by Start-time Fair Queuing– Have virtual time, v(t)

– Each put gets a start time S(pci) and finish time F(pc

i)

F(pci) = S(pc

i) + size(pci) ttl(pc

i)

S(pci) = max(v(A(pc

i)) - , F(pci-1))

v(t) = maximum start time of all accepted puts

Fall 2006 CS 561 39

Fairness with Different Arrival Times

Fall 2006 CS 561 40

Fairness With Different Sizes and TTLs

Fall 2006 CS 561 41

Performance

• Only 28 of 7 million values lost in 3 months– Where “lost” means unavailable for a full hour

• On Feb. 7, 2005, lost 60/190 nodes in 15 minutes to PL kernel bug, only lost one value

Fall 2006 CS 561 42

Performance

• Median get latency ~250 ms– Median RTT between hosts ~ 140 ms

• But 95th percentile get latency is atrocious– And even median spikes up from time to time

Fall 2006 CS 561 43

The Problem: Slow Nodes

• Some PlanetLab nodes are just really slow– But set of slow nodes changes over time

– Can’t “cherry pick” a set of fast nodes

– Seems to be the case on RON as well

– May even be true for managed clusters (MapReduce)

• Modified OpenDHT to be robust to such slowness– Combination of delay-aware routing and redundancy

– Median now 66 ms, 99th percentile is 320 ms • using 2X redundancy