Supporting Scalable and Cooperative Interval Caching in a Clustered Video Server

Abstract

The design of scalable video servers with high throughput is receiving considerable attention with the growing popularity of VOD services. This paper presents a design solution we call scalable and cooperative interval caching (SCIC) for clustered video servers. SCIC uses interval caching and cache cooperation done by hash-based request distribution, and a randomized way of probing for an available cache. This way, SCIC can provide high scalability and throughput in clustered servers with a reasonably small amount of communication among hosts. Our simulation results demonstrate that SCIC is an effective caching strategy providing significant performance gains in clustered video servers.