Benchmarking Preparation for and Aggregate and Sorting Retrievals in the Multi-Backend Database System

摘要

The scope of this thesis is twofold. The first is to provide a methodology for the performance evaluation of the Multi-Backend Database System, MBDS. The second is to describe the implementation and integration for two new database operations, the aggregate retrieval and the sorted retrieval. The thesis provides the essential tools for the successful evaluation of MBDS. The performance evaluation of MBDS is necessary to validate the performance gains in terms of response-time reduction, and capacity growth in terms of response-time invariance. The implementation and integration of the aggregate retrieval and sorted retrieval provide two advanced data retrieval operations to MBDS. The aggregate retrieval operation allows the user to obtain extremely useful data not inherently available in the data itself. The sorted retrieval operation allows the user to retrieve data and have it presented in a more meaningful fashion. (Author)