The Research And Implementation Of Physical Database Schema Recommender Framework

Physical database design is a key phase in devising and deploying a database system to improve overall system performance. However, the appearance of more and more complicated physical design features make the design procedure very complicated.Database systems and their workloads are become more and more complex and dynamic, and the costs of management are become more and more expensive. Automated tools for physical database design can help reduce the total cost of ownership (TCO) of databases by reducing the DBA's burden in determining the appropriate physical design.The performance of an enterprise database system can depend crucially on its physical database design. Automated tools for physical database design can help reduce the total cost of ownership (TCO) of databases by reducing the DBA's burden in determining the appropriate physical design.In this dissertation, A GPDSR (General Physical Database Schema Recommender) framework is provided to support automatic physical database design for relation database systems. The framework divides automatic physical database design into two phases, the workload mining phase and the design advising phase, including three collaborative agents, Workload Analysis Component (WAC), Feature-Space Construction Component (FCC), and Optimal-Feature Search Component (OSC). An index advisor for OSCAR relational engine is prototyped, and experiment results show its feasibility and efficiency.In the first chapter, we introduce the concepts of database and database management system, and the context of the research..In the second chapter, We review related studies on automatic physical database design and their limitations.In the third chapter, we introduce the GAPDD framework.In the fourth chapter, we introduce a prototyped index advisor for OSCAR database engine with some advising results for database schemas and workloads from generally used database benchmarks.In the fifth chapter, we conclude our work by pointing out the contributions and faultiness of OSIA and future work to perfect it.