An efficient data replication approach for structured peer-to-peer systems

Abstract

The peer-to-peer (P2P) communication model has gained an enormous research interest in the last few years due to its promising properties (e.g., decentralization, self-organization, scalability, etc.) to build large scale distributed applications. However, the P2P systems are characterized by their dynamic behavior as peers can frequently join and leave the network. This property makes the data availability in a such dynamic behavior a crucial problem. Indeed, to provide a high data availability under peer churn, one possible solution is to replicate the same bloc of data into a set of nodes according to a well defined strategy. In this paper, we have deeply investigated the main existing approaches of data replication that have been proposed in the literature. Then, we have proposed an efficient data replication mechanism that overcomes the limitations of these approaches. In order to evaluate these strategies, we have conducted an extensive simulation study. Simulations results show that our strategy performs well compared with others approaches not only in a static environment but also under churn.