Stochastic fluid model for P2P caching evaluation

Abstract

In this paper we propose a stochastic fluid model to analyze the performance of Squirrel, a P2P cooperative Web cache. This work provides a scalable and insightful extension of our previous analysis of Squirrel. This new model provides a closed-form expression for the hit probability when documents are equally popular. Realistic object popularity is also addressed through a clustering approximation. The accuracy of this model is validated by a comparison with discrete-event simulations. Our model allows us to study the impact of various parameters on the performance of the Squirrel system. In particular, we emphasize the importance of taking object popularity into account. We also investigate the impact of clients announcing their departure on the resulting hit probability.