International Business Machines Corporation

John Francis Edwards

Michael Steven Faunce

Brian Robert Muras

Randy Lynn Egan

Robert Douglas Driesch, Jr

Curtis Neal Boger

A method generates at least one statistic for a temporary table during optimization of a database query before the temporary table is created. By doing so, optimization can proceed on a database query irrespective of the fact that the temporary table may not be materialized until the database query is ultimately executed.

Generating statistics for temporary tables during query optimization

A metadata manager is used in a database management system to collect and maintain metadata associated with a database. Multiple optimizer instances are permitted to access the metadata maintained by the metadata manager, often eliminating the need for individual optimizer instances to retrieve and process metadata directly from the database. As such, the overhead associated with operations such as collecting, refining, retrieving and/or maintaining of metadata can be off-loaded from the optimizer instances, often accelerating individual cost estimation calculations by optimizer instances, facilitating reuse of metadata calculations and refinements, and improving metadata consistency between multiple related cost estimates.

Metadata manager for database query optimizer

A database query optimizer processes an expression in a database query, and generates therefrom an operand list and a corresponding truth table that may be represented by a list of binary characters, where the operand list and corresponding truth table represent a disjunct normal form for the expression. Each expression is stored once it is processed into its operand list and corresponding list of binary characters. New queries are processed into component expressions, and each expression is checked to see if the expression was previously processed and stored as a processed expression. If so, the operand list and list of binary characters for the previously-stored expression may be used in processing the current expression. If there is no previously-stored expression that corresponds to the current expression, the previously-stored expressions are checked to see if any correspond to a complement of the current expression. If so, a new expression is easily constructed for the current expression by retrieving the list of binary characters that correspond to the complement expression, and inverting the bits in the list of binary characters. If there is no previously-stored expression that corresponds to the current expression or its complement, an operand list and corresponding list of binary characters are generated for the current expression. Logical operations between predicates in a query may be performed by performing mathematical operations on the lists of binary characters corresponding to each predicate expression. The end result is an operand list and corresponding list of binary characters that represents the entire expression in a query.

Growing community of inventors

Rochester, MN, United States of America

John Francis Edwards