An Optimizer for Heterogeneous Systems with NonStandard Data and Search Capabilities

Much of the world's nonstandard data resides in specialized data sources. This data must often be queried together with data from other sources to give users the information they desire. Queries that can span several specialized sources with diverse search capabilities pose new challenges for query optimization. This paper describes the design and illustrates the use of a general purpose optimizer that can be taught about the capabilities of a new data source. Our optimizer produces plans of high quality even in the presence of nonstandard data, strange methods, and unusual query processing capabilities, and makes it easy to add new sources of standard or nonstandard data.