US 12,105,713 B2
Compressing workloads for scalable index tuning
Tarique Ashraf Siddiqui, Redmond, WA (US); Saehan Jo, Ithaca, NY (US); Wentao Wu, Bellevue, WA (US); Chi Wang, Redmond, WA (US); Vivek Ravindranath Narasayya, Redmond, WA (US); and Surajit Chaudhuri, Kirkland, WA (US)
Assigned to MICROSOFT TECHNOLOGY LICENSING, LLC, Redmond, WA (US)
Filed by Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed on May 10, 2022, as Appl. No. 17/740,660.
Prior Publication US 2023/0367771 A1, Nov. 16, 2023
Int. Cl. G06F 16/00 (2019.01); G06F 11/34 (2006.01); G06F 16/21 (2019.01); G06F 16/22 (2019.01); G06F 16/2453 (2019.01); G06F 16/248 (2019.01)
CPC G06F 16/24549 (2019.01) [G06F 11/3409 (2013.01); G06F 16/21 (2019.01); G06F 16/221 (2019.01); G06F 16/24539 (2019.01); G06F 16/248 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving a workload with a plurality of queries;
computing, for each query of the plurality of queries, query features and a utility;
selecting a query to add to a query subset based on a benefit determined by using the query features and the utility;
determining whether a size of the query subset is less than a query subset size;
updating the query features and the utility for remaining queries of the plurality of queries in response to determining the size of the query subset is less than the query subset size;
selecting another query to add to the query subset based on an updated benefit using updated query features and updated utility in response to determining the size of the query subset is less than the query subset size, wherein the query features of the another query has reduced similarity to the query features of queries within the query subset;
continuing to update the query features and the utility for the remaining queries of the plurality of queries and selecting other queries to add to the query subset based on the updated benefit determined by using the updated query features and the updated utility until the size of the query subset is equal to the query subset size, wherein the query features of the other queries have a reduced similarity to the query features of queries within the query subset; and
adding query weights to selected queries in the query subset in response to determining the size of the query subset is equal to the query subset size.