Building Very Large Overlay NetworksJorg Liebeherr

University of Virginia

HyperCast Project

• HyperCast is a set of protocols for large-scale overlay multicasting and peer-to-peer networking

Research problems:• Investigate which topologies are best

suited for

… a large number of peers… that spuriously join and leave … in a network that is dynamically changing

• Make it easier to write applications for overlays easier by

… providing suitable abstractions (“overlay socket”)

… shielding application programmer from overlay topology issues

Acknowledgements

• Team: – Past: Bhupinder Sethi, Tyler

Beam, Burton Filstrup, Mike Nahas, Dongwen Wang, Konrad Lorincz, Jean Ablutz, Haiyong Wang, Weisheng Si, Huafeng Lu

– Current: Jianping Wang, Guimin Zhang, Guangyu Dong, Josh Zaritsky

• This work is supported in part by the National Science Foundation:

D E N A L I

Applications with many receivers

Numberof Receivers

Number of Senders

Streaming

SoftwareDistribution

10 1,000

1

1,000,000

10

1,000

1,000,000

CollaborationTools

Games

Peer-to-PeerApplications

Sensor networks

Need for Multicasting ?

• Maintaining unicast connections is not feasible• Infrastructure or services needs to support a “send to group”

Multicast support in the network infrastructure (IP Multicast)

• Reality Check:

– Deployment has encountered severe scalability limitations in both the size and number of groups that can be supported

– IP Multicast is still plagued with concerns pertaining to scalability, network management, deployment and support for error, flow and congestion control

Overlay Multicasting

• Logical overlay resides on top of the layer-3 network infrastructure (“underlay”)

• Data is forwarded between neighbors in the overlay• Transparent to underlay network• Overlay topology should match the layer-3 infrastructure

Overlay-based approaches for multicasting

• Build an overlay mesh network and embed trees into the mesh:– Narada (CMU)– RMX/Gossamer (UCB)– more

• Build a shared tree:– Yallcast/Yoid (NTT, ACIRI)– AMRoute (Telcordia, UMD – College Park)– Overcast (MIT)– Nice (UMD)– more

• Build an overlay using distributed hash tables and embed trees:– Chord (UCB, MIT)– CAN (UCB, ACIRI)– Pastry/Scribe (Rice, MSR)– more

HyperCast Overlay Topologies

Applications organize themselves to form a logical overlay network with a given topology

Data is forwarded along the edges of the overlay network

Hypercube

Delaunaytriangulation

Spanning tree(for mobile ad hoc)

Programming in HyperCast: Overlay Socket

• Socket-based API

• UDP (unicast or multicast) or TCP

• Supports different semantics for transport of data

• Supports different overlay protocols

• Implementation in Java

OverlaySocket

Forwarding Engine Message Store

Overlay Socket Interface

Sta

tis

tic

s I

nte

rfa

ce

Messages ofthe OverlayProtocol

ApplicationReceiveBuffer

Overlay Node

Overlay Node Interface

Node Adapter

Adapter Interface

Socket Adapter

Adapter Interface

ApplicationMessages

Application Program

Underlay Network

• HyperCast Project website: http://www.cs.virginia.edu/hypercast

Network of overlay sockets

• Overlay socket is an endpoint of communication in an overlay network

• An overlay network is a collection of overlay sockets

• Overlay sockets in the same overlay network have a common set of attributes

• Each overlay network has an overlay ID, which can be used the access attributes of overlay

• Nodes exchange data with neighbors in the overlay topology

Internet

Overlaysocket

Application

Overlaysocket

Application

ApplicationOverlaysocket

Application

ApplicationOverlaysocket

Application

Unicast and Multicast in overlays

Root(sender)

Root(receiver)

Multicast Unicast

Sender

• Unicast and multicast is done along trees that are embedded in the overlay network.

• Requirement: Overlay node must be able to compute the child nodes and parent node with respect to a given root

Overlay Message Format

1 2 3 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 +-------+---------------+---+---+-------------------------------+ |Version|LAS|Dmd| Traffic Class | Flow Label | Next Header | +-------+---------------+---+---+-------------------------------+ | OL Message Length | Hop Limit | +-------------------------------+-------------------------------+ | Src LA | +--------------------------------------------------------------- | Dest LA | +---------------------------------------------------------------+

• Common Header:

Version (4 bit): Version of the protocol (current Version is 0x0)LAS (2 bit): Size of logical address fieldDmd (4bit) Delivery Mode (Multicast, Flood, Unicast, Anycast)Traffic Class (8 bit): Specifies Quality of Service class (default: 0x00)Flow Label (8 bit): Flow identifierNext Header (8 bit) Specifies the type of the next header following this header OL Message Length (8 bit) Specifies the type of the next header following this header.Hop Limit (16 bit): TTL field Src LA ((LAS+1)*4 bytes) Logical address of the source Dest LA ((LAS+1)*4 bytes Logical address of the destination

Loosely modeled after IPv6 minimal header with extensions

Socket Based API

//Generate the configuration object OverlayManager om = new OverlayManager(“hypercast.prop”); String overlayID = om.getDefaultProperty(“MyOverlayID")OverlaySocketConfig config = new om.getOverlaySocketConfig(overlayID);

//create an overlay socket OL_Socket socket = config.createOverlaySocket(callback);

//Join an overlay socket.joinGroup();

//Create a message OL_Message msg = socket.createMessage(byte[] data, int length);

//Send the message to all members in overlay network socket.sendToAll(msg);

//Receive a message from the socket OL_Message msg = socket.receive();

//Extract the payload byte[] data = msg.getPayload();

//Generate the configuration object OverlayManager om = new OverlayManager(“hypercast.prop”); String overlayID = om.getDefaultProperty(“MyOverlayID")OverlaySocketConfig config = new om.getOverlaySocketConfig(overlayID);

//create an overlay socket OL_Socket socket = config.createOverlaySocket(callback);

//Join an overlay socket.joinGroup();

//Create a message OL_Message msg = socket.createMessage(byte[] data, int length);

//Send the message to all members in overlay network socket.sendToAll(msg);

//Receive a message from the socket OL_Message msg = socket.receive();

//Extract the payload byte[] data = msg.getPayload();

• Tries to stay close to Socket API for UDP Multicast

• Program is independent of overlay topology

Property File

# This is the Hypercast Configuration File##  # LOG FILE:LogFileName = stderr # ERROR FILE:ErrorFileName = stderr # OVERLAY Server: OverlayServer =

# OVERLAY ID:OverlayID = 224.228.19.78/9472KeyAttributes = Socket,Node,SocketAdapter # SOCKET:Socket = HCast2-0HCAST2-0.TTL = 255HCAST2-0.ReceiveBufferSize = 200HCAST2-0.ReadTimeout = 0

. . .

# This is the Hypercast Configuration File##  # LOG FILE:LogFileName = stderr # ERROR FILE:ErrorFileName = stderr # OVERLAY Server: OverlayServer =

# OVERLAY ID:OverlayID = 224.228.19.78/9472KeyAttributes = Socket,Node,SocketAdapter # SOCKET:Socket = HCast2-0HCAST2-0.TTL = 255HCAST2-0.ReceiveBufferSize = 200HCAST2-0.ReadTimeout = 0

. . .

# SOCKET ADAPTER:SocketAdapter = TCP SocketAdapter.TCP.MaximumPacketLength = 16384SocketAdapter.UDP.MessageBufferSize = 100

# NODE:Node = HC2-0HC2-0.SleepTime = 400HC2-0.MaxAge = 5HC2-0.MaxMissingNeighbor = 10HC2-0.MaxSuppressJoinBeacon = 3 

# NODE ADAPTER:#NodeAdapter = UDPMulticast NodeAdapter.UDP.MaximumPacketLength = 8192NodeAdapter.UDP.MessageBufferSize = 18NodeAdapter.UDPServer.UdpServer0 = 128.143.71.50:8081NodeAdapter.UDPServer.MaxTransmissionTime = 1000NodeAdapter.UDPMulticastAddress = 224.242.224.243/2424

# SOCKET ADAPTER:SocketAdapter = TCP SocketAdapter.TCP.MaximumPacketLength = 16384SocketAdapter.UDP.MessageBufferSize = 100

# NODE:Node = HC2-0HC2-0.SleepTime = 400HC2-0.MaxAge = 5HC2-0.MaxMissingNeighbor = 10HC2-0.MaxSuppressJoinBeacon = 3 

# NODE ADAPTER:#NodeAdapter = UDPMulticast NodeAdapter.UDP.MaximumPacketLength = 8192NodeAdapter.UDP.MessageBufferSize = 18NodeAdapter.UDPServer.UdpServer0 = 128.143.71.50:8081NodeAdapter.UDPServer.MaxTransmissionTime = 1000NodeAdapter.UDPMulticastAddress = 224.242.224.243/2424

• Stores attributes that configure the overlay socket (overlay protocol, transport protocol and addresses)

Hypercast Software: Demo Applications

Distributed Whiteboard Multicast file transfer

Data aggregation in P2P Net: Homework assignment in CS757 (Computer Networks)

Application: Emergency Response Network for Arlington County, Virginia

Project directed by B. Horowitz and S. Patek

Application of P2P Technology to

Advanced Emergency Response Systems

Priority Messaging

Situational Awareness

Dynamic Grouping andInformation Management

Other Features of Overlay Socket

Current version (2.0):• Statistics: Each part of the overlay socket maintains statistics which are

accessed using a common interface • Monitor and control: XML based protocol for remote access of statistics

and remote control of experiments• LoToS: Simulator for testing and visualization of overlay protocols• Overlay Manager: Server for storing and downloading overlay

configurations

Next version (parts are done, release in Summer 2004):– “MessageStore”: Enhanced data services, e.g., end-to-end reliability,

persistence, streaming, etc.– HyperCast for mobile ad-hoc networks on handheld devices– Service differentiation for multi-stream delivery– Clustering– Integrity and privacy with distributed key management

Delaunay Triangulation Overlays

0,0 1,0 2,0 3,0

0,1

0,2

0,3

12,0

10,8

5,2

4,9

0,6

Nodes are assigned x-y coordinates

(e.g., based on geographic location)

Nodes are assigned x-y coordinates

(e.g., based on geographic location)

Nodes in a Plane

12,0

10,8

5,2

4,9

0,6

The Voronoi region of a node is the region of the plane that is closer to this node than to any other node.

The Voronoi region of a node is the region of the plane that is closer to this node than to any other node.

Voronoi Regions

The Delaunay triangulation has edges between nodes in neighboring Voronoi regions.

The Delaunay triangulation has edges between nodes in neighboring Voronoi regions.

12,0

10,8

5,2

4,9

0,6

Delaunay Triangulation

An equivalent definition:A triangulation such that each circumscribing circle of a triangle formed by three vertices, no vertex of is in the interior of the circle.

An equivalent definition:A triangulation such that each circumscribing circle of a triangle formed by three vertices, no vertex of is in the interior of the circle.

12,0

10,8

5,2

4,9

0,6

Delaunay Triangulation

A

B

C

D

Locally Equiangular Property

• Sibson 1977: Maximize the minimum angle

For every convex quadrilateral formed by triangles ACB and ABD that share a common edge AB, the minimum internal angle of triangles ACB and ABD is at least as large as the minimum internal angle of triangles ACD

and CBD.

A

B

C

D

Next-hop routing with Compass Routing

• A node’s parent in a spanning tree is its neighbor which forms the smallest angle with the root.

• A node need only know information on its neighbors – no routing protocol is needed for the overlay.

Root Node

B

A

15°

30°

B is the Node’s Parent

12,0

4,9

Spanning tree when node (8,4) is root. The tree can be calculated by both parents and children.

Spanning tree when node (8,4) is root. The tree can be calculated by both parents and children.

0,6

5,2

4,9

12,0

10,8

8,4

Evaluation of Delaunay Triangulation overlays

• Delaunay triangulation can consider location of nodes in an (x,y) plane, but is not aware of the network topology

Question: How does Delaunay triangulation compare with other overlay topologies?

Overlay Topologies

Delaunay Triangulation and variants– DT– Hierarchical DT– Multipoint DT

Hypercube

Degree-6 Graph– Similar to graphs generated in Narada

Degree-3 Tree– Similar to graphs generated in Yoid

Logical MST– Minimum Spanning Tree

overlays that assume knowledge of network topology

overlays used by HyperCast

Transit-Stub Network

Transit-Stub• GeorgiaTech

topology generator

• 4 transit domains

• 416 stub domains

• 1024 total routers

• 128 hosts on stub domain

Evaluation of Overlays

• Simulation:– Network with 1024 routers (“Transit-Stub” topology)– 2 - 512 hosts

• Performance measures for trees embedded in an overlay network:

– Degree of a node in an embedded tree

– “Stretch”: Ratio of delay in overlay to shortest path delay

– “Stress”: Number of duplicate transmissions over a physical link

Illustration of Stress and Stretch

AA

BBStress = 2

Stress = 2

Stretch for AB: 1.5

1 1

1 1

Unicast delay AB : 4

11 1

1

1

1

Delay AB in overlay: 6

Average StretchS

tret

ch (

90th P

erce

ntil

e)

90th Percentile of Stretch

Delaunay triangulation

Str

etch

(90

th P

erce

ntil

e)

Average Stress

Delaunay triangulation

90th Percentile of Stress

Delaunay triangulation

The DT Protocol

Protocol which organizes members of a network in a Delaunay Triangulation

• Each member only knows its neighbors

• “soft-state” protocol

Topics:• Nodes and Neighbors• Example: A node joins• State Diagram• Rendezvous• Measurement Experiments

Each node sends Hello messages to its neighbors periodically

Each node sends Hello messages to its neighbors periodically

12,0

5,2

4,9

0,6

10,8HelloHelloH

ello

HelloHello

Hello

HelloHello

HelloHelloH

ello

Hello

HelloHello

• Each Hello contains the clockwise (CW) and counterclockwise (CCW) neighbors

• Receiver of a Hello runs a “Neighbor test” ( locally equiangular prop.)

• CW and CCW are used to detect new neighbors

• Each Hello contains the clockwise (CW) and counterclockwise (CCW) neighbors

• Receiver of a Hello runs a “Neighbor test” ( locally equiangular prop.)

• CW and CCW are used to detect new neighbors

12,0

5,2

4,9

0,6

10,8

Hello

CW =

12,

0

CCW =

4,9

Nei

gh

bo

r

5,2 12,0 4,9 4,9 5,2 –12,0 – 10,8

CC

W

CW

Neighborhood Table of 10,8

A node that wants to join the triangulation contacts a node that is “close”

A node that wants to join the triangulation contacts a node that is “close”

12,0

10,8

5,2

4,9

0,6

8,4

New node

Hello

Node (5,2) updates its Voronoi region, and the triangulation

Node (5,2) updates its Voronoi region, and the triangulation

12,0

4,9

0,6

8,4

10,8

5,2

(5,2) sends a Hello which contains info for contacting its clockwise and counterclockwise neighbors

(5,2) sends a Hello which contains info for contacting its clockwise and counterclockwise neighbors

Hello

12,0

4,9

0,6

8,4

10,8

5,2

12,0

4,9

0,6

(8,4) contacts these neighbors ...(8,4) contacts these neighbors ...

8,4

10,8

5,2

Hello

Hello

12,0

4,9

12,0

4,9

… which update their respective Voronoi regions.

… which update their respective Voronoi regions.

0,6

10,8

5,2

4,9

12,0

8,4

12,0

4,9

0,6

(4,9) and (12,0) send Hellos and provide info for contacting their respective clockwise and counterclockwise neighbors.

(4,9) and (12,0) send Hellos and provide info for contacting their respective clockwise and counterclockwise neighbors.

10,8

5,2

4,9

12,0

8,4

Hello

Hello

12,0

4,9

0,6

(8,4) contacts the new neighbor (10,8) ...

(8,4) contacts the new neighbor (10,8) ...

10,8

5,2

4,9

12,0

8,4H

ello

10,8

12,0

4,9

0,6

…which updates its Voronoi region...…which updates its Voronoi region...

5,2

4,9

12,0

10,8

8,4

12,0

4,9

0,6

…and responds with a Hello…and responds with a Hello

5,2

4,9

12,0

10,8

8,4H

ello

12,0

4,9

This completes the update of the Voronoi regions and the Delaunay Triangulation

This completes the update of the Voronoi regions and the Delaunay Triangulation

0,6

5,2

4,9

12,0

10,8

8,4

Rendezvous Methods

• Rendezvous Problems:– How does a new node detect a member of the overlay?– How does the overlay repair a partition?

• Three solutions:

1. Announcement via broadcast

2. Use of a rendezvous server

3. Use `likely’ members (“Buddy List”)

12,0

10,8

5,2

4,9

0,6

8,4

New node

Hello

Rendezvous Method 1: Announcement via broadcast (e.g., using IP Multicast)

Rendezvous Method 1: Announcement via broadcast (e.g., using IP Multicast)

12,0

10,8

5,2

4,9

0,6

Leader Rendezvous Method 1:

A Leader is a node with a Y-coordinate higher than any of its neighbors.

Rendezvous Method 1:

A Leader is a node with a Y-coordinate higher than any of its neighbors.

12,0

10,8

5,2

4,9

0,6

8,4

New node

Rendezvous Method 2: New node and leader contact a rendezvous server. Server keeps a cache of some other nodes

Rendezvous Method 2: New node and leader contact a rendezvous server. Server keeps a cache of some other nodes

Server

ServerRequestServerReply(12,0)

NewNode

Hello NewNode

12,0

10,8

5,2

4,9

0,6

8,4

New nodewith Buddy List: (12,0) (4,9)

Rendezvous Method 3: Each node has a list of “likely” members of the overlay network

Rendezvous Method 3: Each node has a list of “likely” members of the overlay network

NewNode

Hello NewNode

State Diagram of a Node

Leaderwithout

Neighbor

Leader withNeighbor

NotLeader

Leaving

Stopped

Neighbor added(with smaller coordinates)

All neighborsleave or timeout

Neighbor added(with larger coordinates)

All neighborsleave or timeout

Application starts

A new neighbor withgreater coordinates is added

After removing some neighbor,this node has largest coordinates

Send Goodbye Send Goodbye

SendGoodbye

Applicationexits

Sub-states of a Node

Stable WithoutCandidateNeighbor

Stable WithCandidateNeighbor

NotStable

Node contained in NewNodepasses neighbor test

After handling thecandidate neighbor,node remains stable

After neighborhoodupdating, node becomes

not stable

After neighborhood updating,node becomes stable.

After neighborhoodupdating,

node becomes not stable

• A node is stable when all nodes that appear in the CW and CCW neighbor columns of the neighborhood table also appear in the neighbor column

• A node is stable when all nodes that appear in the CW and CCW neighbor columns of the neighborhood table also appear in the neighbor column

Measurement Experiments

• Experimental Platform: Centurion cluster at UVA (cluster of 300 Linux PCs) – 2 to 10,000 overlay members – 1–100 members per PC

• Random coordinate assignments

Switch 8

Switch 9

Switch 11

Switch 10

Switch 4

Switch 5

Switch 6

Switch 7

Switch 3

Internet

centurion149-167centurion183centurion253-255

centurion246centurion250centurion251

centurion249centurion252

centurion168-182centurion164-187

centurion188-211

centurion228-247centurion128-147

Gigabit Ethernet

How long does it take to add M members to an overlay network of N members ?

Experiment: Adding Members

M+N members

Tim

e to

Co

mp

lete

(se

c)

Experiment: Throughput of Multicasting

Number of Members N

Bandwidth bounds(due to stress)

Measuredvalues

Ave

rag

e th

rou

gh

pu

t

(Mb

ps)

100 MB bulk transfer for N=2-100 members (1 node per PC) 10 MB bulk transfer for N=20-1000 members (10 nodes per PC)

Experiment: Delay

100 MB bulk transfer for N=2-100 members (1 node per PC) 10 MB bulk transfer for N=20-1000 members (10 nodes per PC)

Del

ay o

f a

pac

ket

(mse

c)

Number of Nodes N

Wide-area experiments

• PlanetLab: worldwide network of Linux PCs• Obtain coordinates via “triangulation”

– Delay measurements to 3 PCs

– Global Network Positioning (Eugene Ng, CMU)

Planetlab Video Streaming Demo

Summary

• Overlay socket is general API for programming P2P systems• Several proof-of-concept applications • Intensive experimental testing:

– Local PC cluster (on 100 PCs)• 10,000 node overlay in < 1 minute

• Throughput of 2 Mbps for 1,000 receivers

– PlanetLab (on 60 machines)• Up to 2,000 nodes

• Video streaming experiment with 80 receivers at 800 kbps had few losses

HyperCast (Version 2.0) site: http://www.cs.virginia.edu/hypercastDesign documents, download software, user manual