Parallel Query Processing In Data Warehouse Management System

In recent years, the research on Data Warehouse is becoming the hotspot in database field. A lot of research work has been done domestically and abroad. Many corporations developed their Data Warehouse products, such as the MetaCube of INFORMIX corporation, Discover and Express of ORACLE corporation. But, these products are based on the relational database management system, and there is no flexible query language to satisfy the need of OLAP, especially no language designed aiming at the multi-dimensionality of data warehouse. A Parallel Data Warehouse management system(PDWMS) working in parallel computing environment has been developed by our group. It is an independent Data Warehouse management system, adopting a multi-dimensional model, having various parallel multi-dimensional storage structure, and providing a query language DWSQL which can do all kinds of OLAP operations. The focus of this dissertation is on the research of the PDWMS syntax analyzer, query processor and pre-compiler. In this dissertation, we propose the syntax analyzing and semantic checking algorithms, query decomposing strategy and algorithms for different operations, and introduce how to pre-compile the C program which includes the sentences of DQSWL and to transform the sentences of DWSQL into the functions which the compiler of C language can recognize, and how to transfer parameters between C language and PDWMS.