P2PEACE: a P2P-based simulation Environment for Autonomic Content Exchange Networks

This paper presents a study on a global replication architecture intended to autonomic content exchange between communities of active nodes. The main objective is to exploit and reuse the synergy of groups to optimize data access in a distributed system. A self management based on thematic similarity between nodes is an efficient manner of optimizing data access (availability and selectivity). The evaluation and the test of such architecture before its large deployment are, however, extremely important. Therefore, a simulation tool called P2PEACE is developed. It is designed primarily to visualize interactions between the network nodes and provide some related statistics to measure the exchange performance. In fact, the existing simulation environments offer good tools of testing the performance of an isolated node (e.g. Web cache) or algorithms used for cooperation between groups. However, these approaches do not give a clear vision of the content mobility between the system nodes. In this paper, a method to analyze collective nodes as a global system and understand how contents move around them is, therefore, proposed. Results showed that a semantic organization can greatly optimize the exchanges in such a situation