Performance analysis of a P2P storage system with a lazy replica repair policy
Shunfu Jin, Yuan Zhao, Wuyi Yue, Lingling Chen · Journal of Industrial and Management Optimization · 2013
Peer-to-Peer (P2P) storage systems are a prevalent and importantmode for implementing cost-efficient, large-scale distributedstorage. Considering the random departure feature of the peers andthe diverse popularity of the data objects, a proper number ofreplicas needs to be maintained, and a reasonable trigger thresholdof replica repair needs to be set for high data availability andlow system overhead. In this paper, based on the working principleof the lazy replica repair policy in a P2P storage system, athree-dimensional Markov chain model is constructed, and the modelis analyzed in steady-state by using a matrix-geometric method.Then, the performance measures in terms of the availability of onedata object, the average access latency, and the replication rateare given. Moreover, numerical results with analysis are providedto demonstrate how system parameters such as the replica number andthe replica repair instant influence the system performance.Finally, we develop benefit functions to optimize the replica numberand the repair trigger threshold.