The rapid growth of the Web has sparked significant research activity on improving the performance and scalability of the Web. One obvious approach to improving performance is caching: Web proxy caches or multi-level caching hierarchies can be used to store copies of popular Web objects closer to clients, thereby reducing network and server load, and providing faster response times to Web users.

This talk describes the design and use of a synthetic Web proxy workload generator to evaluate the effectiveness of Web proxy caching architectures. The synthetic workload generator can be used to closely match the salient characteristics of empirical Web proxy workloads (e.g., one-timers, Zipf-like referencing behaviour, heavy-tailed file size distribution, temporal locality), or be used to generate multiple workloads that differ from each other in carefully controlled ways (e.g., Zipf slope, tail index, correlation between document popularity and document size).

We use the workload generator, and trace-driven simulation, to address two main research questions: "How sensitive are Web proxy caching algorithms to workload characteristics?" and "How can workload characteristics be exploited to improve the performance of multi-level Web caching hierarchies?".

The simulation results highlight the sensitivity of Web caching performance to Zipf slope, temporal locality, and the correlation (if any) between document popularity and document size, but the relative insensitivity to one-timers and the heavy tail index. Several novel multi-level caching algorithms are also investigated, and their performance tradeoffs discussed.

Carey Williamson is a Professor in the Department of Computer Science at the University of Saskatchewan. He received a B.Sc.(Honours) in Computer Science from the University of Saskatchewan in 1985, and a Ph.D. in Computer Science from Stanford University in 1991. His general research interests are in computer networks and computer systems performance evaluation. More specific interests include network traffic measurement, workload characterization, network simulation, and Web performance. Dr. Williamson is a member of IEEE, ACM, and SCS.