Businesses today rely on large collections of data stored in diverse systems with differing capabilities. Database middleware systems offer the possibility of inter-relating data from these various sources via a single high-level query interface. One key issue for these systems is how to process queries in an environment in which each source of data may have its own unique set of query processing capabilities. In this paper we present the design of a query opotimizer for Garlic, a middleware system designed to integrate data from a broad range of data sources, with very different query capabilities. Garlic's optimizer extends the rule-based approach of Lohman to work in a heterogeneous environment, by allowing multiple sets of rules, each representing the capabilities of a particular data source. This approach offers great advantages in terms of plan quality, extensibility to new sources, incremental immplementation of rules for new sources, and the ability to express the capabilities of a diverse set of sources. We describe the design and implementation of this optimizer, and present results of experiments that show the importance of doing optimization in this environment.
By: Donald Kossmann (Univ. of Passau, Germany), Laura M. Haas, Edward L. Wimmers and Jun Yang (Stanford Univ.)
Published in: RJ10065 in 1997
This Research Report is not available electronically. Please request a copy from the contact listed below. IBM employees should contact ITIRC for a copy.
Questions about this service can be mailed to reports@us.ibm.com .