Multi-criteria Index Selection for Grouped SQL Queries

Abstract

Indexing is a key element of optimization of relational database systems (RDBMS). Commercial tools supporting index selection (e.g. SQL Access Advisor, Toad, SQL Server Database Tuning Advisor, DB2 Advisor) are based on the methods dedicated to individual queries. This paper presents a new approach to tables indexing for the SQL queries group that take into account the size of the indexes and their creation time. Examples illustrate that the use of the group concept reduces the query execution time by 44,% compared to classical methods. © Springer-Verlag Berlin Heidelberg 2013.