Amir Rasti Daniel Stutzbach Reza Rejaie The ION P2P Project University of Oregon On the Long-term Evolution of the Two-Tier.

Slides:

Advertisements

Similar presentations

Characterizing Overlay Topologies & Dynamics in Peer-to-Peer Networks Daniel Stutzbach, Reza Rejaie University of Oregon Subhabrata Sen AT&T Labs IEEE.

Advertisements

Peer-to-Peer and Social Networks An overview of Gnutella.

1 Greedy Forwarding in Dynamic Scale-Free Networks Embedded in Hyperbolic Metric Spaces Dmitri Krioukov CAIDA/UCSD Joint work with F. Papadopoulos, M.

Clayton Sullivan PEER-TO-PEER NETWORKS. INTRODUCTION What is a Peer-To-Peer Network A Peer Application Overlay Network Network Architecture and System.

Walter Willinger AT&T Research Labs Reza Rejaie, Mojtaba Torkjazi, Masoud Valafar University of Oregon Mauro Maggioni Duke University HotMetrics’09, Seattle.

Amir Rasti Reza Rejaie Dept. of Computer Science University of Oregon.

1 An Overview of Gnutella. 2 History The Gnutella network is a fully distributed alternative to the centralized Napster. Initial popularity of the network.

Search and Replication in Unstructured Peer-to-Peer Networks Pei Cao, Christine Lv., Edith Cohen, Kai Li and Scott Shenker ICS 2002.

LightFlood: An Optimal Flooding Scheme for File Search in Unstructured P2P Systems Song Jiang, Lei Guo, and Xiaodong Zhang College of William and Mary.

Denial-of-Service Resilience in Peer-to-Peer Systems D. Dumitriu, E. Knightly, A. Kuzmanovic, I. Stoica and W. Zwaenepoel Presenter: Yan Gao.

Topology Generation Suat Mercan. 2 Outline Motivation Topology Characterization Levels of Topology Modeling Techniques Types of Topology Generators.

Ten Years in the Evolution of the Internet Ecosystem

Small-World Graphs for High Performance Networking Reem Alshahrani Kent State University.

Small-world Overlay P2P Network

Measurement and Estimation of Network QoS among Peer Xbox Game Players Youngki Lee, KAIST Sharad Agarwal, Microsoft Research Chris Butcher, Bungie Studio.

A Framework for Architecting Peer-to- Peer Receiver-driven Overlays Reza Rejaie, Shad Stafford Mirage Research Group Department of Computer Science University.

Improving Lookup Performance over a Widely-Deployed DHT Daniel Stutzbach Reza Rejaie The ION P2P Project University of.

 We developed a fast and tunable crawler, Cruiser.  Cruiser uses a master-slave architecture, parallel crawling, and leverages the two-tier topology.

Characterizing the Two-Tier Gnutella Topology  Gnutella, FastTrack, and eDonkey use two-tier overlay topologies.  Our initial study focuses on Gnutella.

Efficient Content Location Using Interest-based Locality in Peer-to-Peer Systems Presented by: Lin Wing Kai.

Issues in Offering Live P2P Streaming Service to Residential Users Nazanin Magharei, *Yang Guo, and Reza Rejaie Dept. of Computer and Information Science.

Exploiting Content Localities for Efficient Search in P2P Systems Lei Guo 1 Song Jiang 2 Li Xiao 3 and Xiaodong Zhang 1 1 College of William and Mary,

Count / Top-k Continuous Queries on P2P Networks 01/11/2006.

Understanding Mesh-based Peer-to-Peer Streaming Nazanin Magharei Reza Rejaie.

Understanding Churn in Peer-to-Peer Networks Daniel Stutzbach – University of Oregon Reza Rejaie – University of Oregon Internet Measurement Conference.

1 Characterizing Files in the Modern Gnutella Network: A Measurement Study Shanyu Zhao, Daniel Stutzbach, Reza Rejaie University of Oregon SPIE Multimedia.

Chord-over-Chord Overlay Sudhindra Rao Ph.D Qualifier Exam Department of ECECS.

Graphs and Topology Yao Zhao. Background of Graph A graph is a pair G =(V,E) –Undirected graph and directed graph –Weighted graph and unweighted graph.

Characterizing Unstructured Overlay Topologies in Modern P2P File-Sharing Systems Daniel Stutzbach – University of Oregon Reza Rejaie – University of Oregon.

1 Characterizing Selfishly Constructed Overlay Routing Networks March 11, 2004 Byung-Gon Chun, Rodrigo Fonseca, Ion Stoica, and John Kubiatowicz University.

On Unbiased Sampling for Unstructured Peer-to-Peer Networks Daniel Stutzbach – University of Oregon Reza Rejaie – University of Oregon Nick Duffield –

A Framework for Cost-Effective Peer-to- Peer Content Distribution Mohamed Hefeeda and Bharat Bhargava Department of Computer Sciences Purdue University.

UNIVERSITY OF JYVÄSKYLÄ Resource Discovery Using NeuroSearch Presentation for the Agora Center InBCT-seminar Mikko Vapa, researcher InBCT 3.2.

UNIVERSITY OF JYVÄSKYLÄ Resource Discovery in Unstructured P2P Networks Distributed Systems Research Seminar on Mikko Vapa, research student.

Allocations vs Announcements A comparison of RIR IPv4 Allocation Records with Global Routing Announcements Geoff Huston May 2004 (Activity supported by.

Analyzing Peer-to-Peer Traffic Across Large Networks Jia Wang Joint work with Subhabrata Sen AT&T Labs - Research.

A Measurement-driven Analysis of Information Propagation in the Flickr Social Network WWW09 报告人：徐波.

Presentation by Manasee Conjeepuram Krishnamoorthy.

(Social) Networks Analysis III Prof. Dr. Daning Hu Department of Informatics University of Zurich Oct 16th, 2012.

Analysis and Modeling of the Open Source Software Community Yongqin Gao, Greg Madey Computer Science & Engineering University of Notre Dame Vincent Freeh.

Introduction Widespread unstructured P2P network

1 Reading Report 4 Yin Chen 26 Feb 2004 Reference: Peer-to-Peer Architecture Case Study: Gnutella Network, Matei Ruoeanu, In Int. Conf. on Peer-to-Peer.

Developing Analytical Framework to Measure Robustness of Peer-to-Peer Networks Niloy Ganguly.

IEEE P2P, Aachen, Germany, September Ad-hoc Limited Scale-Free Models for Unstructured Peer-to-Peer Networks Hasan Guclu

Application-level versus Network-level Proximity Mohammad Malli PhD student Planete Project, INRIA - Sophia Antipolis France.

Introduction to Peer-to-Peer Networks. What is a P2P network A P2P network is a large distributed system. It uses the vast resource of PCs distributed.

Network Aware Resource Allocation in Distributed Clouds.

Understanding Population Dynamics. Agenda Layout 1234 The world at 7 billion Demographic transitions 3 Patterns of population change Strategies needed.

A Measurement-driven Analysis of Information Propagation in the Flickr Social Network author: Meeyoung Cha Alan Mislove Krishna P. Gummadi From Saarbrucken,

A Routing Underlay for Overlay Networks Akihiro Nakao Larry Peterson Andy Bavier SIGCOMM’03 Reviewer: Jing lu.

Multimedia Computing & Networking Shanyu Zhao, Daniel Stutzbach, Reza Rejaie Multimedia & Internetworking Research Group (Mirage) Computer & Information.

Super-peer Network. Motivation: Search in P2P Centralised (Napster) Flooding (Gnutella)  Essentially a breadth-first search using TTLs Distributed Hash.

Quantitative Evaluation of Unstructured Peer-to-Peer Architectures Fabrício Benevenuto José Ismael Jr. Jussara M. Almeida Department of Computer Science.

Selfishness, Altruism and Message Spreading in Mobile Social Networks September 2012 In-Seok Kang

Freelib: A Self-sustainable Digital Library for Education Community Ashraf Amrou, Kurt Maly, Mohammad Zubair Computer Science Dept., Old Dominion University.

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.

Content Distribution in Unstructured Peer-to-Peer Networks Daniel Stutzbach Committee Members: Professor Reza Rejaie Professor Ginnie Lo Professor Art.

Efficient P2P Search by Exploiting Localities in Peer Community and Individual Peers A DISC’04 paper Lei Guo 1 Song Jiang 2 Li Xiao 3 and Xiaodong Zhang.

Vulnerability in Socially-informed Peer-to-Peer Systems Jeremy Blackburn Nicolas Kourtellis Adriana Iamnitchi University of South Florida.

P2PComputing/Scalab 1 Gnutella and Freenet Ramaswamy N.Vadivelu Scalab.

"A Measurement Study of Peer-to-Peer File Sharing Systems" Stefan Saroiu, P. Krishna Gummadi Steven D. Gribble, "A Measurement Study of Peer-to-Peer File.

Sampling Techniques for Large, Dynamic Graphs Daniel Stutzbach – University of Oregon Reza Rejaie – University of Oregon Nick Duffield – AT&T Labs—Research.

LightFlood: An Efficient Flooding Scheme for File Search in Unstructured P2P Systems Song Jiang, Lei Guo, and Xiaodong Zhang College of William and Mary.

Plethora: Infrastructure and System Design. Introduction Peer-to-Peer (P2P) networks: –Self-organizing distributed systems –Nodes receive and provide.

Brief Announcement : Measuring Robustness of Superpeer Topologies Niloy Ganguly Department of Computer Science & Engineering Indian Institute of Technology,

Mapping the Gnutella Network: Properties of Large-Scale Peer-to-Peer Systems and Implications for System Design Authors: Matei Ripeanu Ian Foster Adriana.

1 Finding Spread Blockers in Dynamic Networks (SNAKDD08)Habiba, Yintao Yu, Tanya Y., Berger-Wolf, Jared Saia Speaker: Hsu, Yu-wen Advisor: Dr. Koh, Jia-Ling.

Large-Scale Monitoring of DHT Traffic Ghulam Memon – University of Oregon Reza Rejaie – University of Oregon Yang Guo – Corporate Research, Thomson Daniel.

Department of Computer Science University of York

Joydeep Chandra, Santosh Shaw and Niloy Ganguly

Presentation transcript:

Amir Rasti Daniel Stutzbach Reza Rejaie The ION P2P Project University of Oregon On the Long-term Evolution of the Two-Tier Gnutella Overlay Global Internet Symposium Barcelona, Spain April 28 th, 2006

Daniel Stutzbach The ION P2P Project 2/16 Peer-to-peer applications are becoming increasingly popular. How do they adapt to long-term growth? Case study: The Gnutella population quadrupled in the 15 months between Oct 04 and Jan 06, from 800k to 3.2M. How does this affect properties of the network? Motivation

Daniel Stutzbach The ION P2P Project 3/16 Gnutella uses a two-tier overlay. Improves scalability. Ultrapeers form an unstructured mesh. Leaf peers connect to the ultrapeers. Leaf peers self-promote to ultrapeers when they can’t find an available ultrapeer. Studying the overlay requires snapshots. Snapshots capture the overlay as a graph. Individual snapshots reveal graph properties. Consecutive snapshots reveal dynamics. Background Top-level overlay Leaf Ultrapeer

Daniel Stutzbach The ION P2P Project 4/16 Cruiser, a P2P topology crawler Daniel Stutzbach and Reza Rejaie, “Capturing Accurate Snapshots of the Gnutella Network”, the Global Internet Symposium, Contacts only ultrapeers. Captures 2.2M peers in around 8 minutes. For each ultrapeer, records: Brand, version, and list of neighbors Characterizing the Gnutella topology Daniel Stutzbach, Reza Rejaie, and Subhabrata Sen, “Characterizing Unstructured Overlay Topologies in Modern P2P File-Sharing Systems”, Internet Measurement Conference, Examined various graph and dynamic properties Looked for similarities across different snapshots Prior Work

Daniel Stutzbach The ION P2P Project 5/16 Outline Looking for differences in snapshots Client properties Brand Version Graph properties Percentage of ultrapeers Node degree Pair-wise distances Resiliency Geographic Properties Location Correlations with topology

Daniel Stutzbach The ION P2P Project 6/16 Data Set To examine long-term trends, we need a series of comparable snapshots. We have 20,000 snapshots from Oct through Jan Unfortunately they’re not evenly distributed. Many of them are back-to-back. We selected 18 comparable snapshots taken at 3pm on weekdays.

Daniel Stutzbach The ION P2P Project 7/16 Population and Brand Population has grown linearly. 15%—20% Ultrapeers LimeWire and BearShare have a fairly steady marketshare. Dip in LimeWire users in winter 2004—2005.

Daniel Stutzbach The ION P2P Project 8/16 Version We breakdown LimeWire users by version. Users upgrade quickly 50% running latest version within 2 months of release. New features spread quickly. “Flag Days” feasible.

Daniel Stutzbach The ION P2P Project 9/16 Percentage of Ultrapeers Each leaf aims for 3 parents and each ultrapeer may have up to 30 children. If everyone is full, 9% of peers would be ultrapeers. If there are too few ultrapeers, some leaves are stuck! If there are too many, efficiency decreases. In practice, we want slightly higher than 9%. Peaks at 18% and 20% show unnecessarily high number of ultrapeers. Decreases correlated with release of new versions of LimeWire.

Daniel Stutzbach The ION P2P Project 10/16 Node Degree: Ultrapeer-to-Ultrapeer LimeWire ultrapeers try to maintain around 30 neighbors. We observe an increasing number of peers with 20— 30 neighbors. The number of peers with fewer neighbors has not increased much. Conclusion: Ultrapeers scalably find one another.

Daniel Stutzbach The ION P2P Project 11/16 Node Degree: Ultrapeer-to-Leaf Number of children peak at 30 and 45, corresponding with the maximums used in LimeWire and BearShare. The peaks have hardly changed. The number of ultrapeers with many open slots has increased. Conclusion: Leaves are having an increasingly hard time finding ultrapeers, and are unnecessarily self-promoting. However, this trend is reversed in the latest snapshots (2006)

Daniel Stutzbach The ION P2P Project 12/16 Pairwise Distance Distribution of distances between pairs of peers Important for searching Despite a dramatic change in population, only a slight change in distances. Similar to behavior in random graphs

Daniel Stutzbach The ION P2P Project 13/16 Resiliency How many peers must be removed before the network is severely fragmented? Random Remove peers at random More than 90% must be removed. Little change Systematic Remove high degree peers first Typically more than 60% must be removed. Dip in winter 2004—2005, then little change

Daniel Stutzbach The ION P2P Project 14/16 Geographic Location We examined the geographic location of ultrapeers. 98.5% of all users are in: North America (NA), Europe (EU), Asia (AS), or South America (SA). Relatively little change in geographic distribution over time. NA has decreased slightly, while EU has increased.

Daniel Stutzbach The ION P2P Project 15/16 Geographical-Topological Correlation We find that a peer’s neighbors are disproportionately from within the same region. The effect is more pronounced in regions with fewer users. Causes: Locale-based neighbor selection Latency as a natural selection criteria Time of day effect

Daniel Stutzbach The ION P2P Project 16/16 Conclusions We examined trends in the properties of a two- tier overlay, Gnutella, over 15 months. Gnutella witnessed a major increase in user- population over this time. The percentage of ultrapeers tends to creep upwards over time. Rapid adaptation of new releases enables developers to modify the software to cope with the growing user population. Peers tend to connect to other peers in the same geographic region.

Daniel Stutzbach The ION P2P Project 17/16 Clustering Coefficient (CC) CC is inverseley related to the percentage of ultrapeers. Degree distribution in top layer is almost fixed so more nodes in top layer makes the graph sparser -> lower CC.

Daniel Stutzbach The ION P2P Project 18/16 Node Degree: Leaf-to-Ultrapeer Peak in 1-3 parents growing with population increase. No other significant change.