Design And Implementation Of Memory Database Storage Management

Because of that the memory can stored all or most of the data of Main Memory Database (MMDB), there is little I/O during transaction processing of MMDB. So, which is different from traditional Disk-Resident Database (DRDB), reducing I/O is no more the goal of optimization of MMDB. The different means that the various methods for DRDB are no longer suitable for MMDB, it is necessary to propose new methods for processing data in MMDB based on the features of itself. As the foundation of MMDB, storage management is very important. Storage management influences not only the performance significantly, but also the high-level functions of MMDB. So it is very meaningful to study the storage management of MMDB.Based-on the above requirements, this thesis studies the storage management in MMDB, and presents the design and implementation of the corresponding storage manager. Specifically, the study includes data organization, memory management, and index mechanism. The contributions of the thesis mainly include:presents the storage hierarchy of the storage manager and the concrete organization of records based-on the analysis of storage hierarchy in MMDB; improves the performance of the storage manager by improving the format of records; designs and implements a memory management method by using multiple memory pools, the method can efficiently utilizes the memory; designs a flexible startup load policy and a data exchange policy; deeply analyze the features and usability of the hash index, T-tree index and cache-conscious index mechanism, and presents the details of design, implementation and optimization methods of T-tree index.At the last, the experiments show that the size of index node has influence on performance of index, and the improvement of record format significantly improve the performance of the storage manager. And the experiment contrasted to FastDB also shows that the performance of the storage manager is equivalent to or higher than the performance of FastDB, which means that performance of the storage manager is feasible in practice.