SCRIBE A large-scale and decentralized application-level multicast infrastructure.

Slides:



Advertisements
Similar presentations
Dynamic Replica Placement for Scalable Content Delivery Yan Chen, Randy H. Katz, John D. Kubiatowicz {yanchen, randy, EECS Department.
Advertisements

Ion Stoica, Robert Morris, David Karger, M. Frans Kaashoek, Hari Balakrishnan MIT and Berkeley presented by Daniel Figueiredo Chord: A Scalable Peer-to-peer.
Pastry Peter Druschel, Rice University Antony Rowstron, Microsoft Research UK Some slides are borrowed from the original presentation by the authors.
Peter Druschel, Rice University Antony Rowstron, Microsoft Research UK
Impact Analysis of Cheating in Application Level Multicast s 1090176 Masayuki Higuchi.
Kademlia: A Peer-to-peer Information System Based on the XOR Metric Petar Mayamounkov David Mazières A few slides are taken from the authors’ original.
SplitStream by Mikkel Hesselager Blanné Erik K. Aarslew-Jensen.
SplitStream: High- Bandwidth Multicast in Cooperative Environments Monica Tudora.
Pastry Peter Druschel, Rice University Antony Rowstron, Microsoft Research UK Some slides are borrowed from the original presentation by the authors.
1 PASTRY Partially borrowed from Gabi Kliot ’ s presentation.
Common approach 1. Define space: assign random ID (160-bit) to each node and key 2. Define a metric topology in this space,  that is, the space of keys.
Scribe An application level multicast infrastructure Kasper Egdø and Morten Bjerre.
Pastry: Scalable, decentralized object location and routing for large-scale peer-to-peer systems Antony Bowstron & Peter Druschel Presented by: Long Zhang.
Scribe: A Large-Scale and Decentralized Application-Level Multicast Infrastructure Miguel Castro, Peter Druschel, Anne-Marie Kermarrec, and Antony L. T.
ZIGZAG A Peer-to-Peer Architecture for Media Streaming By Duc A. Tran, Kien A. Hua and Tai T. Do Appear on “Journal On Selected Areas in Communications,
SplitStream: High-Bandwidth Multicast in Cooperative Environments Marco Barreno Peer-to-peer systems 9/22/2003.
Applications over P2P Structured Overlays Antonino Virgillito.
Pastry: Scalable, decentralized object location and routing for large-scale peer-to-peer systems Antony Rowstron and Peter Druschel Proc. of the 18th IFIP/ACM.
Scalable Application Layer Multicast Suman Banerjee Bobby Bhattacharjee Christopher Kommareddy ACM SIGCOMM Computer Communication Review, Proceedings of.
Spring 2003CS 4611 Peer-to-Peer Networks Outline Survey Self-organizing overlay network File system on top of P2P network Contributions from Peter Druschel.
June, 2002INFOCOM 1 Host Multicast: A Framework for Delivering Multicast to End Users Beichuan Zhang (UCLA) Sugih Jamin (UMich) Lixia Zhang (UCLA)
SCRIBE: A large-scale and decentralized application-level multicast infrastructure Miguel Castro, Peter Druschel, Anne-Marie Kermarrec and Antony Rowstron.
Mobile and Wireless Computing Institute for Computer Science, University of Freiburg Western Australian Interactive Virtual Environments Centre (IVEC)
SCALLOP A Scalable and Load-Balanced Peer- to-Peer Lookup Protocol for High- Performance Distributed System Jerry Chou, Tai-Yi Huang & Kuang-Li Huang Embedded.
1 An Overlay Scheme for Streaming Media Distribution Using Minimum Spanning Tree Properties Journal of Internet Technology Volume 5(2004) No.4 Reporter.
Application Layer Multicast for Earthquake Early Warning Systems Valentina Bonsi - April 22, 2008.
Wide-area cooperative storage with CFS
An Evaluation of Scalable Application-level Multicast Using Peer-to-peer Overlays Miguel Castro, Michael B. Jones, Anne-Marie Kermarrec, Antony Rowstron,
1 Peer-to-Peer Networks Outline Survey Self-organizing overlay network File system on top of P2P network Contributions from Peter Druschel.
P2P Course, Structured systems 1 Introduction (26/10/05)
Pastry: Scalable, decentralized object location and routing for large-scale peer-to-peer systems (Antony Rowstron and Peter Druschel) Shariq Rizvi First.
Mobile Ad-hoc Pastry (MADPastry) Niloy Ganguly. Problem of normal DHT in MANET No co-relation between overlay logical hop and physical hop – Low bandwidth,
Tapestry GTK Devaroy (07CS1012) Kintali Bala Kishan (07CS1024) G Rahul (07CS3009)
1 PASTRY. 2 Pastry paper “ Pastry: Scalable, decentralized object location and routing for large- scale peer-to-peer systems ” by Antony Rowstron (Microsoft.
PIC: Practical Internet Coordinates for Distance Estimation Manuel Costa joint work with Miguel Castro, Ant Rowstron, Peter Key Microsoft Research Cambridge.
ON DESIGING END-USER MULTICAST FOR MULTIPLE VIDEO SOURCES Y.Nakamura, H.Yamaguchi, A.Hiromori, K.Yasumoto †, T.Higashino and K.Taniguchi Osaka University.
Overcast: Reliable Multicasting with an Overlay Network CS294 Paul Burstein 9/15/2003.
Information-Centric Networks07a-1 Week 7 / Paper 1 Internet Indirection Infrastructure –Ion Stoica, Daniel Adkins, Shelley Zhuang, Scott Shenker, Sonesh.
Multicast Routing Algorithms n Multicast routing n Flooding and Spanning Tree n Forward Shortest Path algorithm n Reversed Path Forwarding (RPF) algorithms.
Chapter 22 Network Layer: Delivery, Forwarding, and Routing Part 5 Multicasting protocol.
Content Addressable Network CAN. The CAN is essentially a distributed Internet-scale hash table that maps file names to their location in the network.
An IP Address Based Caching Scheme for Peer-to-Peer Networks Ronaldo Alves Ferreira Joint work with Ananth Grama and Suresh Jagannathan Department of Computer.
P2p file storage and distribution Team: Brian Smith, Daniel Suskin, Dylan Nunley, Forrest Vines Mentor: Brendan Burns.
November 17, 2015Department of Computer Sciences, UT Austin1 SDIMS: A Scalable Distributed Information Management System Praveen Yalagandula Mike Dahlin.
Peer to Peer A Survey and comparison of peer-to-peer overlay network schemes And so on… Chulhyun Park
Pastry: Scalable, decentralized object location and routing for large-scale peer-to-peer systems Antony Rowstron and Peter Druschel, Middleware 2001.
Information-Centric Networks10b-1 Week 10 / Paper 2 Hermes: a distributed event-based middleware architecture –P.R. Pietzuch, J.M. Bacon –ICDCS 2002 Workshops.
Application Layer Multicast Extensions to RELOAD draft-kolberg-sam-baseline-protocol-00 ALM Usage for RELOAD John Buford, Mario Kolberg 3/29/2011.
Information-Centric Networks Section # 10.2: Publish/Subscribe Instructor: George Xylomenos Department: Informatics.
Peer to Peer Network Design Discovery and Routing algorithms
CS 6401 Overlay Networks Outline Overlay networks overview Routing overlays Resilient Overlay Networks Content Distribution Networks.
Peter R Pietzuch and Jean Bacon Peer-to-Peer Overlay Networks in an Event-Based Middleware DEBS’03, San Diego, CA, USA,
Click to edit Master title style Multi-Destination Routing and the Design of Peer-to-Peer Overlays Authors John Buford Panasonic Princeton Lab, USA. Alan.
LOOKING UP DATA IN P2P SYSTEMS Hari Balakrishnan M. Frans Kaashoek David Karger Robert Morris Ion Stoica MIT LCS.
1 Protocol Independent Multicast (PIM) To develop a scalable protocol independent of any particular unicast protocol –ANY unicast protocol to provide routing.
Scribe: A Large-Scale and Decentralized Application-Level Multicast Infrastructure Miguel Castro, Peter Druschel, Anne-Marie Kermarrec, and Antony I.T.
1 FairOM: Enforcing Proportional Contributions among Peers in Internet-Scale Distributed Systems Yijun Lu †, Hong Jiang †, and Dan Feng * † University.
Peer-to-Peer Networks 05 Pastry Christian Schindelhauer Technical Faculty Computer-Networks and Telematics University of Freiburg.
Group Communication A group is a collection of users sharing some common interest.Group-based activities are steadily increasing. There are many types.
Fabián E. Bustamante, Fall 2005 A brief introduction to Pastry Based on: A. Rowstron and P. Druschel, Pastry: Scalable, decentralized object location and.
Chapter 29 Peer-to-Peer Paradigm Copyright © The McGraw-Hill Companies, Inc. Permission required for reproduction or display.
MZR: A Multicast Protocol based on Zone Routing
Pastry Scalable, decentralized object locations and routing for large p2p systems.
COS 461: Computer Networks
Host Multicast: A Framework for Delivering Multicast to End Users
CS5412: Using Gossip to Build Overlay Networks
CS5412: Using Gossip to Build Overlay Networks
COS 461: Computer Networks
Kademlia: A Peer-to-peer Information System Based on the XOR Metric
CS5412: Using Gossip to Build Overlay Networks
Presentation transcript:

SCRIBE A large-scale and decentralized application-level multicast infrastructure

Overview  Pastry  PAST  distributed file system layered on top of Pastry  SCRIBE  decentralized publish/subscribe system

Pastry – Quick Review  Chord like routing  Consistent hashing  Prefix routing  Leaf set

Pastry – locality properties  Short routes  Total distance traveled  Average dist 1.59 to 2.2 times actual dist  Route convergence  Dist Traveled by 2 messages sent to same key  Equal to dist between to nodes before routes converge

Pastry API  nodeID = pastryInit(Credentials)  Causes node to join pastry network  route(msg,key)  send(msg,IP-addr)  Applications must export:  deliver(msg,key)  forward(msg,key,nextID)  newLeafs(leafSet)

SCRIBE  Built on top of Pastry  Support large number of groups  Handle a high rate of membership turnover  SCRIBE nodes can:  Create groups  Join groups  Multicast messages to groups

SCRIBE API  create(credentials, groupID)  join(credentials, groupID)  leave(credentials, groupID)  multicast(credentials, groupID, message)

SCRIBE – Creating a Group  Pastry route(msg, key)  SCRIBE route(CREATE, groupID)  groupID => hash textual name cat creator name  Message delivered to closest key which become rendez-vous point for the group (root of multicast tree for group)  Adds to local list of groups  Stores credentials  Alternative use itself as root  good choice if creator sends to group often

SCRIBE – Joining a Group  Pastry route(msg, key)  SCRIBE route(JOIN, groupID)  routed to rendez-vous point  along the way multicast tree formed

SCRIBE – Leaving a Group  Remove from local group children list  If list becomes empty forward to parent  Part of the multicast tree may be removed

SCRIBE – Sending a multicast message  route(MULTICAST, groupID)  ask for rendez-vous IP address  If rendez-vous fails re-request rendez-vous point  Pastry handles node duplication  All messages are sent through the rendez-vous point

SCRIBE – Repairing the Multicast Tree  Messages are delivered only in best-effort  may be out of order delivery  Periodic heartbeat message sent to all children  Child rejoins the tree through sending a new JOIN message if suspects parent has failed  Can repair rendez-vous point  Pastry handles node duplication in leaf nodes  Children nodes JOIN new root when missing heartbeat is detected

SCRIBE – Forming a Multicast Tree  Rendez-vous point (root)  Forwarders  may or may not be members of the group  maintain a children table (IP and nodeID) for group

SCRIBE - Strengths  Pastry handles root duplication  Rendez-vous point does not handle all join requests  Locality properties of Pastry  short routes delay from rendez-vous point to member is short  route convergence load imposed on physical network is small

SCRIBE – Experimental Evalutation  Simulation experimental results  Focus on three metrics:  delay to deliver events to group members  stress on each node  stress on each physical network link

SCRIBE – Simulator Evaluation  5050 routers and 100,000 end nodes  1,500 groups of different sizes  10 different runs using same parameters but different random seeds  Averaged all results  Compared results with IP multicast

SCRIBE – Delay Penalty  RMD – ratio between max delay using SCRIBE & max delay using IP multicast  RAD – ratio between average delay using SCRIBE  & average delay using IP multicast

SCRIBE – Node Stress  Average node is responsible for forwarding a small number of multicast messages

SCRIBE – Link Stress  Total num links = 1,035,295  SCRIBE = 2,489,824 messages (mean 2.4)  IP multicast = 758,853 messages (mean 0.7)

SCRIBE – Bottleneck remover  Bottlenecks  Low capacity nodes  High capacity nodes with extremely high children entries  Drop children if over capacity  Select child to drop and send message with children table  Child chooses new parent node and sends JOIN message  Result  Removes long tail in node stress graph  Increases average link stress

SCRIBE – Scalability Small Groups  50,000 nodes  30,000 groups 11 members each group  SCRIBE performs poorly for large number small groups  SCRIBE collapse  Removes long paths removing nodes that are not members of a group & have only one entry in their children table

SCRIBE – Scalability Small Groups  Average link stress 6.1 to 3.3  Average number of children 21.2 to 8.5

SCRIBE - Conclusion  Fully decentralized  Support large number of groups  Support large group size  Multiple multicast sources per group  QUESTIONS