A Multi-Query Optimizer for Monet

Database systems allow for concurrent use of several applications (and query interfaces). Each application generates an ``optimal'' plan---a sequence of low-level database operators---for accessing the database. The queries posed by users through the same application can be optimized together using traditional multi-query optimization techniques. However, the commonalities among queries of different applications are not exploited. In this paper we present an efficient inter-application multi-query optimizer that re-uses previously computed (intermediate) results and eliminates redundant work. Experimental results on a single CPU system and a parallel system show that the inter-application multi-query optimizer improves the query evaluation performance significantly.