Aging through cascaded caches: Performance issues in the distribution of Web content

The Web is a distributed system, where data is stored and disseminated from both origin servers and caches. Origin servers provide the most up-to-date copy whereas caches store and serve copies that had been cached for a whil e. Origin servers do not maintain per-client state, and weak-consistency of cached copies is maintained by the origin server attaching to each copy an expiration time. Typically, the lifetime-duration of an object is fixed, a nd as a result, a copy fetched directly from its origin server has maximum time-to-live (TTL) whereas a copy obtained through a cache has a shorter TT L since its age (elapsed time since fetched from the origin) is deducted fr om its lifetime duration. Thus, a cache that is served from a cache would i ncur a higher miss-rate than a cache served from origin servers. Similarly, a high-level cache would receive more requests from the same client popula tion than an origin server would have received. As Web caches are often ser ved from other caches (e.g., proxy and reverse-proxy caches), age emerges a s a performance factor. Guided by a formal model and analysis, we use diffe rent inter-request time distributions and trace-based simulations to explor e the effect of age for different cache settings and configurations. We als o evaluate the effectiveness of frequent pre-term refreshes by higher-level caches as a means to decrease client misses. Beyond Web content distributi on, our conclusions generally apply to systems of caches deploying expirati on-based consistency.