US Patent Application 17897930. SCALABLE INDEX TUNING WITH INDEX FILTERING AND INDEX COST MODELS simplified abstract

SCALABLE INDEX TUNING WITH INDEX FILTERING AND INDEX COST MODELS

Organization Name

Microsoft Technology Licensing, LLC==Inventor(s)==

[[Category:Tarique Ashraf Siddiqui of Redmond WA (US)]]

[[Category:Vivek Ravindranath Narasayya of Redmond WA (US)]]

[[Category:Surajit Chaudhuri of Kirkland WA (US)]]

[[Category:Wentao Wu of Kirkland WA (US)]]

SCALABLE INDEX TUNING WITH INDEX FILTERING AND INDEX COST MODELS - A simplified explanation of the abstract

This abstract first appeared for US patent application 17897930 titled 'SCALABLE INDEX TUNING WITH INDEX FILTERING AND INDEX COST MODELS

Simplified Explanation

The patent application describes a method for training an index filter in an index tuning system. This system aims to improve the performance of databases by optimizing the use of indexes.

• The method involves receiving different workloads and databases, each containing various tables and queries.
• Labeled training data is generated by making optimizer calls to a query optimizer using query and index configuration pairs from the databases and workloads.
• An index filter model is trained to identify signals in the labeled training data that indicate potential performance improvements when using specific index configurations for queries.
• The index filter model is also trained to learn rules for identifying spurious indexes, which are indexes that do not provide any performance benefits.
• The trained index filter model is stored in memory for use in the index tuning system.

Original Abstract Submitted

A method of training an index filter for an index tuning system includes receiving a plurality of different workloads and a plurality of different databases, each database including different tables and each workload including a plurality of queries; generating labeled training by making optimizer calls to a query optimizer using query and index configuration pairs from the plurality of databases and the plurality of workloads; training an index filter model to identify signals in the labeled training data, the signals being indicative of a potential performance improvement associated with using an index configuration for a given query; training the index filter model to learn rules over the signals for identifying spurious indexes; and storing the index filter model in a memory.