A dynamic interquery optimization method for achieving data sharing among concurrent queries

A method, called dynamic interquery optimization, for identifying and sharing the intermediate/final results produced by the relational algebraic operations of concurrent database queries is presented. This method determines if any portion of a new query will produce results that are identical to, or are a subset of, the results that have already been produced or are being produced by any executing query operation. Query tree matching and merging algorithms are also given. The computational overhead of the proposed method is analyzed, and cost formulas for calculating the matching time required for establishing different types of matches are provided. The complexity of matching and merging of query trees is compared with the execution-time savings resulting from the elimination of redundant operations. The results verify the benefits of intermediate data sharing.<
>