locality aware network solutions

37
Locality Aware Network Solutions Dahlia Malkhi The Hebrew University of Jerusalem

Upload: terri

Post on 07-Jan-2016

36 views

Category:

Documents


0 download

DESCRIPTION

Locality Aware Network Solutions. Dahlia Malkhi The Hebrew University of Jerusalem. A Brief Overview of Distributed Computing. The 90 ’ s: Internet activity: Web browsing Paradigm: Client-server Techniques: cluster computing, Paxos, group communication. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Locality Aware Network Solutions

Locality Aware Network Solutions

Dahlia MalkhiThe Hebrew University of Jerusalem

Page 3: Locality Aware Network Solutions

3

A Brief Overview of Distributed Computing

The 90’s: 2000-

– Internet activity: File sharing

– Paradigm: P2P, grid, web-services

– Techniques: overlay networks, content distribution networks, resource location

Page 4: Locality Aware Network Solutions

4

Application: IPv6 Routing over IPv4[van Renesse 02]

AF3S:::3FF1:43E4

0001:::3BBB:5555

1111:::7777:7754

2222:::2222:2222

EEE0:::EEEE:EEEE

5151:::6161:6666

567A:::0202:0202

8888:::0909:9999

Distribute Hash Tables (DHT)

Page 5: Locality Aware Network Solutions

5

Application: Content Delivery / Finding Nearest Copies of Data

?

?

?

Page 6: Locality Aware Network Solutions

6

Application: Hyperencryption[Maurer 92, Ding & Rabin 02]

Random bits

Alice BobKey

Adversary bits

Page 7: Locality Aware Network Solutions

7

Application: A Hyperencryption P2P Network

[Rabin 03]Distributed Hash Table

(DHT)

Page 8: Locality Aware Network Solutions

8

Application: A Distributed Google?

?

Page 9: Locality Aware Network Solutions

9

Scalable Network Solutions

Overlay networks provide added functionality at the application level– Search, routing, location services

Network theory provides the foundations– Possibilities, impossibilities, lower/upper bounds

Practical solutions require flexible deployment

Page 10: Locality Aware Network Solutions

10

Distributed Data Structures (DDS)

Peers jointly implement a data structure, e.g., hash table

Route queries based on data-name (key)

Page 11: Locality Aware Network Solutions

11

DDS Problem Reduced to Routing?? 00001111

Responsible for 00001111

Page 12: Locality Aware Network Solutions

12

Why classic routing network designs don’t help

Static # of nodes a priori

known Node labels

designated by network designer

000

111

110

101001

100

011

010

Page 13: Locality Aware Network Solutions

13

DDS Reduced to Routing

The problem: Overlay routing network– Variants: labeled routing,

name-independent routing, finding nearest copies

Dynamic emulation

Page 14: Locality Aware Network Solutions

14

Distributed Hash Tables

[Malkhi, Naor, Ratacjzak, PODC 2002]

SchemeDegreeRoute

Chord, Tapestry, Pastry [2001]

Log nLog n

CAN [2001]

dd*n(1/d)

Viceroy [2002]

5Log n

Koorde, D2B, DH, generic

[2003]

2Log n

[Abraham, Awerbuch, Azar, Bartal, Malkhi, Pavlov, IPDPS 2003]

Page 15: Locality Aware Network Solutions

15

Tree View of Dynamic Graphs

Leafs of the tree represent current nodes

Inner nodes in the tree represent nodes that were split

000

111

110

101

00

110 111

001

100

011

010

000 001 010 011 100 101

00

Example: merge of 000, 001 into 00

Page 16: Locality Aware Network Solutions

16

Locality awareness

source

target

Page 17: Locality Aware Network Solutions

17

Locality awareness

source

target

Page 18: Locality Aware Network Solutions

18

Locality Awareness in Overlay Networks

Model the network as a weighted undirected graph– c(x, y): cost of shortest path from x to y– c() is a metric

An overlay network is a sub-graph Let x=x0 , x1, …, xt=y be a route in the overlay

network Stretch: Ratio between overlay route cost and

shortest path cost:( c(x, x1) + c(x1,x2) + … c(xt-1, y) ) / c(x,y)

Page 19: Locality Aware Network Solutions

19

Overlay Networks inGrowth-Bounded Metrics

Previous work:– [Plaxton, Rajaraman, Rica 1997], Tapestry (Berkeley),

Pastry (MS UK)

– Expected (large) constant stretch– Logarithmic node degree

LAND [Abraham, Malkhi, Dobzinski, SODA 2004]:– Guaranteed stretch (1+ε)– Expected logarithmic node degree, constant

depends on growth-bound– Simple, intuitive construction and proofs

Page 20: Locality Aware Network Solutions

20

Overlay Networks in Geometric Spaces

Modeling the Internet as a geometric space is practical– Ubiquitous GPS devices– Successful embeddings in virtual

coordinate-space Problem 1: Locate nodes Problem 2: Route to known coordinates

Page 21: Locality Aware Network Solutions

21

Location Services and Routingin Geometric Spaces

LLS: First fully-locality aware location service [Abraham Dolev Malkhi 2004]

– bounded stretch lookup– bounded stretch update

First constant-degree routing scheme (to known coordinates)[Abraham Malkhi, PODC 2004]

– constant node degree, logarithmic hops, 1+ε stretch

Page 22: Locality Aware Network Solutions

22

Routing in Arbitrary Graphs: Lower and upper bounds

Name-independent routing: node names are independent of routing scheme [Awerbuch, Bar Noy, Linial, Peleg 1989]

Lower bounds: [Gavoille Gengler 2001] – Stretch < 3 O(n) routing information– Stretch < 5 √n routing information

Upper bound: [Abraham, Gavoille, Malkhi, Nisan, Thorup, SPAA 2004]

– stretch-3 routing with O(√n ) routing information– Stretch 3 is indeed attainable!

General upper bound: [Abraham Gavoille, Malkhi, DISC 2004]

– Stretch-k routing with memory O(k2 k√n )

Page 23: Locality Aware Network Solutions

23

ever-growing global scale-free networks, their provisioning, repair and unique functions

EVERGROW

The Vision

ultimate RAID

ultimate GNUTELLA

ultimate GOOGLE

ultimate AKAMAI

infrastructure and new methods and systems devoted to measurement, mock-up and and analysis of present and future network traffic, topology and logical structure, to bridge the gap in theory, protocols and understanding to what the Internet can be in 2025.

An EC project. Coordinators: SICS (Sweden) and HUJI (Jerusalem)

Page 24: Locality Aware Network Solutions

24

Locality-Aware, Robust Overlay for

Information Lookup and Content Delivery Degree O(√n) Locality awareness:

– Formally stretch 3– For far-apart nodes, lower stretch

Mostly two-hop– Whenever full connectivity exists

Flexibility – Estimate √n roughly– Cache information on many vicinity nodes– Store information about any known node of same color

Fault tolerance:– Multiple route choices– Quick repair– Maintain QoS in face of churn

Page 25: Locality Aware Network Solutions

25

Page 26: Locality Aware Network Solutions

26

Large Scale Content Delivery

Initially split the content Then cross-exchange data pieces Solutions build on top of overlay routing

networks

Page 27: Locality Aware Network Solutions

27

FastReplica Cherkasova,Lee -HP Labs - 2003

Phase 1 Source

Clients

Page 28: Locality Aware Network Solutions

28

FastReplica Cherkasova,Lee -HP Labs - 2003

Phase 2 Source

Clients

Page 29: Locality Aware Network Solutions

29

Locality motivation – Tree example

Page 30: Locality Aware Network Solutions

30

Julia algorithm motivation: “Divide and conquer”

First phase

Page 31: Locality Aware Network Solutions

31

2nd phase 2nd phase

Julia algorithm motivation: “Divide and conquer”

Page 32: Locality Aware Network Solutions

32

Network nodes

Page 33: Locality Aware Network Solutions

33

Nodes’ random identifiers

0111010

0011110 1111110

0001000

10010010101001

Page 34: Locality Aware Network Solutions

34

Coloring and Vicinities

Page 35: Locality Aware Network Solutions

35

Coloring and Vicinities

?

?

?

?

Page 36: Locality Aware Network Solutions

36

Stretch 3

?

d

≤ d

≤ 2d

Page 37: Locality Aware Network Solutions

37

The Full Routing Scheme

?

12

345

?a

b

c

d