Tuning: um estudo sobre a otimização de desempenho de sistemas gerenciadores de banco de dados relacionais sob carga de trabalho de suporte a decisão

Abstract

The Relational Database Management Systems are more and more required for handling complex databases and with large data volumes. Consequently present the need more accurate performance. In this context, we stands out the tuning, which consists of fine adjustment to parameters Relational Database Management Systems with the objective of improve their performance. This work seek to realise a study on tuning in Relational Database Management Systems focusing on workload of the type Online Analytical Processing extremely common in databases. To simulate the workload previously mentioned and also to analyze the performance, we used the methodology of the TPC-H benchmarks by using the benchmark Database Test 3. Was evaluated in this work through tuning techniques, the sensitivity the configuration parameters of a Relational Database Management Systems, and their behavior under workload Online Analytical Processing and verifying impacts on their performance.