International Business Machines Corporation

Brian L Baggett

Terence Patrick Purcell

Shuanglin Guo

Andrei Fedorovich Lurie

Thomas Abel Beavin

Oliver Draese

Patrick Dooling Bossman

Shengxi Suo

Martina Simova

Michael A Chang

Allan Benjamin Lebovitz

Ou Jin

Martin Dinh

A DBMS query-optimization module receives a corpus of training data that contains data-access requests, such as SQL queries. Each request specifies data to be accessed but does not specify a query plan that the database should use to generate the requested data. The module identifies, in each received request, parameters, such as join methods and access methods, that can vary among query plans; and variables that cannot be assigned values until the query is actually processed. The system generates a set of queries, each of which implements a distinct query plan, that specify every viable permutation of values of the parameters and of the variables. The generated queries are added to the received corpus, which is forwarded to a machine-learning module in order to train the query-optimizer to select query plans that incur the lowest resource costs when servicing a particular type of query.

Permutation-based machine learning for database query optimization

A method, system and computer program product are provided. A first index scan of records of a database table is performed. A first index used by the first index scan is specified in a DBMS query statement as being ordered by one or more field values of one or more fields of the first index. A second scan of the database table is performed to determine whether each respective record of the second scan is qualified based on one or more fields of the each respective record of the second scan and a predicate. A lookup data structure is built indicating whether each of the respective records of the second scan is qualified. The lookup data structure is used to determine whether a record of the first index scan is qualified when the lookup data structure becomes available.

Helper scan in a database management system

An access plan including multiple indexes is selected for accessing data of a database system. A filtering status of each index within the selected access plan is determined based on information available at execution of the selected access plan. The order of the indexes within the selected access plan is modified based on the filtering status of the indexes, and the selected access plan is executed with the modified order of the indexes. The execution of the selected access plan is interrupted in response to accessing a threshold amount of data and the filtering of the indexes of the selected access plan is estimated. A processing order for the indexes is determined based on the estimated filtering. The execution of the selected access plan resumes based on the determined processing order.

Adaptive multi-index access plan for database queries

Provided are techniques for reduced fixed length sort of variable length columns. A fixed length sort is performed to sort a table having rows for a query, wherein the fixed length sort is based on a sort key comprised of a variable length column. A pre-determined number of the rows are read to identify the variable length column as a candidate for truncation. A length for truncating the variable length column is determined based on database statistics and based on the reading of the pre-determined number of the rows. The variable length column is truncated to the determined length to output a truncated variable length column for the sort key. The rows are sorted on the sort key having the truncated variable length column. Results are generated using the sorted rows and returned for the query.

Reduced fixed length sort of variable length columns

Provided are techniques for a Data-Partitioned Secondary Index (DPSI) partition level join. While using a Data-Partitioned Secondary Index (DPSI) to perform a join of an outer table and an inner table, a different task from multiple tasks is assigned to each partition of the inner table. With each task, a join is performed of the outer table and the assigned partition of the inner table using the DPSI to generate results. The results from each different task are merged.

Growing community of inventors

Renton, WA, United States of America

Brian L Baggett