Cache satellite distribution systems: modeling, analysis, and efficient operation

摘要

Web caches have become an integral component contributing to the improvement of the performance observed by Web clients. Cache satellite distribution systems (CSDSs) have emerged as a technology for feeding the caches with the information clients are expected to request, ahead of time. In such a system, the participating proxies periodically report to a central station about requests received from their clients. The central station selects a collection of Web documents, which are "pushed" via a satellite broadcast to the participating proxies, so that upon a future local request for the documents, they will already reside in the local cache, and will not need to be fetched from the terrestrial network. In this paper, our aim is addressing the issues of how to operate the CSDS, how to design it, and how to estimate its effect. Questions of interest are: 1) what Web documents should be transmitted by the central station and 2) what is the benefit of adding a particular proxy into a CSDS? We offer a model for CSDS that accounts for the request streams addressed to the proxies and which captures the intricate interaction between the proxy caches. Unlike models that are based only on the access frequency of the various documents, this model captures both their frequency and their locality of reference. We provide an analysis that is based on the stochastic properties of the traffic streams that can be derived from HTTP logs, examine it on real traffic, and demonstrate its applicability in selecting a set of proxies into a CSDS.