ARe-SQL: An Online, Automatic and Non-Intrusive Approach for Rewriting SQL Queries

  • Arlino H. M. de Araújo Federal University of Piauí
  • José Maria Monteiro Federal University of Ceará
  • José Antônio F. de Macêdo Federal University of Ceará
  • Júlio A. Tavares University of Fortaleza
  • Angelo Brayner University of Fortaleza
  • Sérgio Lifschitz Pontifical Catholic University of Rio de Janeiro
Keywords: Query processing, Database tuning, Query rewriting

Abstract

Database applications have become very complex, dealing with a huge volume of data and database objects. Concurrently, low query response time and high transaction throughput have emerged as mandatory requirements. Among other possible interventions regarding database performance, SQL query rewriting has been shown efficient. The idea is to rewrite a new SQL statement equivalent to the statement initially formulated, where the new SQL statement  provides performance gains w.r.t. query response time. In this article, we propose  an online, automatic and non-intrusive approach for rewriting SQL queries, denoted ARe-SQL. ARe-SQL rewrites SQL statements, using a set of 11 heuristics. Based on ARe-SQL, two different strategies for SQL-query rewriting, denoted assisted and automatic, were implemented. These strategies were evaluated in three different DBMSs considering three distinct scenarios. The results indicate that both strategies can provide performance gains, which, in some cases, reach 92.5%.

Author Biography

José Maria Monteiro, Federal University of Ceará
Professor Adjunto do Departamento de Computação.
Published
2014-07-13