Using Semantic Information to Improve Query Caching

In this paper, we study how using semantic information improves performance of query caching for dynamic content web sites. We design and implement a series of optimizations for a fully transparent dynamic content cache suitable for any web application including applications with strong transactional requirements such as e-commerce workloads. We optimize the query cache to handle full and partial coverage of query results at the expense of additional processing. Full coverage occurs when a query response is fully contained within a cached response. Partial coverage transforms cache misses that would otherwise occur for insert-caused invalidations into partial hits. In a partial hit, a query result is obtained from merging an existing cached response with one or more residual query results that may need to be computed at the database. We conclude that augmenting transparent query caching with the ability to retrieve partial results from the cache improves performance in terms of hit-rate, throughput and latency