Correctness of Gossip-Based Membership under Message Loss

Due to their simplicity and effectiveness, gossip-based membership protocols have become the method of choice for maintaining partial membership in large P2P systems. A variety of gossip-based membership protocols were proposed. Some were shown to be effective empirically, lacking analytic understanding of their properties. Others were analyzed under simplifying assumptions, such as lossless and delay-less network. It is not clear whether the analysis results hold in dynamic networks where both nodes and network links can fail. In this paper we try to bridge this gap. We first enumerate the desirable properties of a gossip-based membership protocol, such as view uniformity, independence, and load balance. We then propose a simple Send & Forget protocol, and show that even in the presence of message loss, it achieves the desirable properties.

[1]  Jim Gray,et al.  Notes on Data Base Operating Systems , 1978, Advanced Course: Operating Systems.

[2]  A. Frieze,et al.  On the connectivity of randomm-orientable graphs and digraphs , 1982 .

[3]  Nancy A. Lynch,et al.  Impossibility of distributed consensus with one faulty process , 1985, JACM.

[4]  Stefan Savage,et al.  The end-to-end effects of Internet path selection , 1999, SIGCOMM '99.

[5]  Anne-Marie Kermarrec,et al.  SCAMP: Peer-to-Peer Lightweight Membership Service for Large-Scale Group Communication , 2001, Networked Group Communication.

[6]  Idit Keidar,et al.  Evaluating the running time of a communication round over the internet , 2002, PODC '02.

[7]  Edith Cohen,et al.  Search and replication in unstructured peer-to-peer networks , 2002, ICS '02.

[8]  Y. Peres,et al.  Evolving sets, mixing and heat kernel bounds , 2003, math/0305349.

[9]  Anne-Marie Kermarrec,et al.  Lightweight probabilistic broadcast , 2003, TOCS.

[10]  Christos Gkantsidis,et al.  Random walks in peer-to-peer networks , 2004, IEEE INFOCOM 2004.

[11]  Idit Keidar,et al.  Araneola: a scalable reliable multicast system for dynamic environments , 2004, Third IEEE International Symposium on Network Computing and Applications, 2004. (NCA 2004). Proceedings..

[12]  M. V. Steen,et al.  Epidemic-style Monitoring in Large-Scale Sensor Networks , 2005 .

[13]  John E. Hopcroft,et al.  Correctness of a gossip based membership protocol , 2005, PODC '05.

[14]  Christian Schindelhauer,et al.  Peer-to-peer networks based on random transformations of connected regular undirected graphs , 2005, SPAA '05.

[15]  A. Allavena.  On the Correctness of Gossip-based Membership Protocols , 2008 .

[16]  Christian Schindelhauer,et al.  Distributed random digraph transformations for peer-to-peer networks , 2006, SPAA '06.

[17]  Anne-Marie Kermarrec,et al.  Peer counting and sampling in overlay networks: random walk methods , 2006, PODC '06.

[18]  Anne-Marie Kermarrec,et al.  Gossip-based peer sampling , 2007, TOCS.

[19]  Chen Avin,et al.  How to Explore a Fast-Changing World (Cover Time of a Simple Random Walk on Evolving Graphs) , 2008, ICALP.

[20]  Roy Friedman,et al.  RaWMS - Random Walk Based Lightweight Membership Service for Wireless Ad Hoc Networks , 2008, TOCS.

[21]  Anne-Marie Kermarrec,et al.  Bridging the Gap between Population and Gossip-based Protocols , 2008 .

[22]  Idit Keidar,et al.  Correctness of gossip-based membership under message loss , 2009, PODC '09.

[23]  Idit Keidar,et al.  Brahms: byzantine resilient random membership sampling , 2008, PODC '08.