Study And Implementation Of The Heterogeneous Database Access And Integration In CGSP

In recent years, grid computing has been focused by scientific communities as a new emerging technology, and we called it as the Next Generation Internet. Grid can integrate lots of geographically distributed heterogeneous resources, including hardware and software resources, to compose virtual organizations, which are connected by high performance network in a broad range. Grid emphasizes the resources sharing and the collaborative work environment. As grid technology developing, more and more grid-based applications need a grid middleware which can access large, heterogeneous data resources. It's just the background of this paper.This paper proposes a new fundamental architecture of the heterogeneous database access and integration in grid, which is based on the ChinaGrid Support Platform project. It presents the concepts of virtual table, physical table and temporary middle database, designs and implements the SQL interpreter and dispatcher of query statements, handles the large data query by file stream, and extends the data transport component based on the functionalities of OGSA-DAI core framework. Meanwhile, it provides uniform webservice-based interfaces, which can help users to share, query and use resources conveniently. Moreover, it is the most important that it provides distributed joint query spanning many the heterogeneous databases.This article is organized as follows. Firstly, it introduces concepts and technologies about grid as well as the information of grid specifications and statuses. It discriminates the grid computing from the traditional cluster computing and distributed computing such as P2P systems. Secondly, the article introduces ChinaGrid Support Platform project, and illustrates the related modules of CGSP in detail. Lastly, the article proposes a novel and feasible infrastructure for heterogeneous database access and integration, defines relevant concepts, depicts the fundamental architecture and how to implement it. The performance evaluation and the experiment results will be discussed at the end of this paper.