Research On Query&Optimization Mechanism Of Database Middleware

Bridging each Information Isolated Island, Database Middleware is one of the most widely used and maturest middleware. When integrating the heterogeneous data sources, Database Middleware improves the effective of implement of data query by using its built-into optimization and translation. However, with the rapid development of Internet enterprise requirements for distributed processing and Internet information network, Database Middleware is challenged by some problems: large numbers of users, the concurrent query requests, a lot of data source sites and massive data access. These problems can be reduced to such two subjects: the architecture of database middleware and its query optimization mechanism.To solve these problems discussed above, the thesis presents a new Database Middleware solution based on data query optimization mechanism of a weighted query tree (W-Tree). In the global query optimization, decreasing the cost of query performance, reducing the size of data shipping and lowering the time of query response are important objects. Therefore, based on characters of data global query on the condition of distributed environment, an optimization algorithm is presented for the global query: an algorithm of optimization based on a weighed query tree. The algorithm minimizes the cost of operators by using optimization of logical perform placement and the size of data shipping by using optimization of physical perform placement. At the same time, the present Database Middleware Systems is improved to adapt to the requirements of running the algorithm, designing MOWCHA (Middleware Based on a W-Tree and Code Shipping Architecture).At last, this thesis analyzes and evaluates an example to demonstrate the feasibility of the new solution.