A theoretical evaluation of peer-to-peer internal clock synchronization

Synchronized clocks are usually considered as a prerequisite for many distributed applications. Existing solutions mainly deal with this problem in static environments with well defined characteristics and limits. The needs of an emergent class of large-scale peer-to-peer applications that have to operate without any assumptions on the surrounding environment have recently revitalized this research area with the proposals of new solutions characterized by self-organization capabilities and strong adaptability to dynamic settings. This paper reports about the properties of a clock synchronization algorithm for large scale applications. The algorithm implements an internal clock synchronization mechanism which combines the gossip-based paradigm with a nature-inspired approach coming from the coupled oscillators phenomenon. Using a theoretical approach, the paper focuses on the convergence properties of the algorithm, characterizing its synchronization speed (decay factor) the final synchronization point and error.

[1]  Danny Dolev,et al.  Linear Time Byzantine Self-Stabilizing Clock Synchronization , 2003, OPODIS.

[2]  David L. Mills,et al.  Network Time Protocol (version 1) specification and implementation , 1988, RFC.

[3]  Danny Dolev,et al.  Self-Stabilizing Pulse Synchronization Inspired by Biological Pacemaker Networks , 2003, Self-Stabilizing Systems.

[4]  P. M. Melliar-Smith,et al.  Synchronizing clocks in the presence of faults , 1985, JACM.

[5]  Flaviu Cristian,et al.  Lower bounds for convergence function based clock synchronization , 1995, PODC '95.

[6]  Anne-Marie Kermarrec,et al.  The Peer Sampling Service: Experimental Evaluation of Unstructured Gossip-Based Implementations , 2004, Middleware.

[7]  Leslie Lamport,et al.  Time, clocks, and the ordering of events in a distributed system , 1978, CACM.

[8]  K. Arvind,et al.  Probabilistic Clock Synchronization in Distributed Systems , 1994, IEEE Trans. Parallel Distributed Syst..

[9]  Nancy A. Lynch,et al.  A new fault-tolerant algorithm for clock synchronization , 1984, PODC '84.

[10]  Michel Dagenais,et al.  Internal Clock Drift Estimation in Computer Clusters , 2008, J. Comput. Networks Commun..

[11]  Danny Dolev,et al.  Fault-tolerant clock synchronization , 1984, PODC '84.

[12]  Shlomi Dolev,et al.  Possible and Impossible Self-Stabilizing Digital Clock Synchronization in General Graphs , 1997, Real-Time Systems.

[13]  Roberto Baldoni,et al.  Impact ofWAN Channel Behavior on End-to-end Latency of Replication Protocols , 2006, 2006 Sixth European Dependable Computing Conference.

[14]  Roberto Baldoni,et al.  An Adaptive Coupling-Based Algorithm for Internal Clock Synchronization of Large Scale Dynamic Systems , 2007, OTM Conferences.

[15]  Flaviu Cristian,et al.  Clock Synchronization in the Presence of Omission and Performance Faults, and Processor Joins , 1986 .

[16]  Flaviu Cristian.  A probabilistic approach to distributed clock synchronization , 1989, [1989] Proceedings. The 9th International Conference on Distributed Computing Systems.