Optimized caching in systems with heterogeneous client populations

Abstract

An important question in providing on-demand access to large widely shared data files, such as popular video files, is how to effectively use regional (proxy) servers that can store some of the data close to the clients. The proxy caching problem is more complex in the context of continuous media files because of the need to consider bandwidth as well as storage constraints at the proxy servers, and because of the bandwidth sharing possibilities provided by recently proposed multicast delivery techniques. This paper develops new highly efficient analytic models for determining optimal proxy cache content in such environments. Specifically, the new models apply to heterogeneous systems in which the proxy servers have different client workloads and server capabilities. Results from the models provide general insights into caching strategies for such systems, and suggest that it may be useful to employ efficient cost models in actual systems to determine what content should be cached in response to the measured client workload.