Research On OLAP Oriented Distributed Data Base System

With the continuous development of Internet technology,the amount of data generated is also growing,the traditional relational database system has become increasingly unable to meet the needs of these massive data storage and rapid retrieval needs,in order to solve this problem,people have been proposed various distributed database system solutions.Meanwhile,the increasingly fierce competition among enterprises,OLAP technology,as an important part of the enterprise decision support system,its importance has been self-evident.The thesis,based on a company's specific business communication needs,studies the OLAP oriented distributed database system's design and implementation.The thesis mainly conducts its research on the target system from the following areas: the architecture,data manipulation and data definition capabilities of the "OLAP Oriented Distributed Database System".The main research works are as follows:(1)The thesis analyzes the construction program of the distributed database system and parallel computing conception,and then combining the two with specific business needs,designs an architecture based on Shared-Nothing for "OLAP Oriented Distributed Database System",then on the basis of the architecture,studies the approximate processing system of users' requests.(2)The thesis studies the key technologies for "OLAP Oriented Distributed Database System" from data manipulation and data definition function.For data manipulation,distributed storage of data distribution scheme and a distributed query in the associated multi-table queries which is often involved OLAP business would be discussed in the thesis.For distributed storage of data distribution scheme,the thesis studied the random distribution of programs and consistent hashing scheme,and came up with the random algorithm for the design of a random distribution;for distributed query,which associated multi-table queries,this thesis mainly probed into the broadcast strategy and the redistribution strategy in distributed computing environments for associated database tables,and these two strategies are optimized.For data definition function,the thesis mainly studies the process of DDL statement.(3)The thesis designs a modular structure of "OLAP Oriented Distributed Database System" and implements SQL parsing,distributed storage and distributed query capabilities of data manipulation functions,and data definition features.Meanwhile,the function of data distributed storage,data distributed queries,and data definition for processing users' request needed to be done in collaboration by a plurality of nodes in the system,thus this thesis designs and implements a cooperator to coordinate and manage each node.