Design and Implementation of a Global Self-tuning Architecture Anolan Yamilé Milanés and Sérgio Lifschitz   View Paper (PDF)   Return to Projeto e Otimização/Database Design and Optimization Abstract Self-tuning is a feature that enables the automation of DBMS functions. The main goal is to keep a good performance while minimizing the DBA interaction with the system. This work proposes an architecture for DBMS automatic tuning. We consider here tuning as a global issue, given that changes of a single parameter can affect the performance of other operations. Besides describing the architecture details, we discuss its integration within PostgreSQL in a implementation based on software agents. This way we provide a more flexible approach to include automatic tuning features in a DBMS. ©2006 Association for Computing Machinery