Research And Application Of Integrated Data Platform Based On Virtual Data Center

With the development of network, a rapidly growing number of applications need to access and manipulate data from multiple databases, so a way that not only could integrate data from multiple databases, but also could provide a unified user interface to access all the databases becomes more and more pressing. However the various data sources distributed heterogeneous environment have different data formats, storage modes, access control, data models, manipulation languages and data semantics. Meanwhile, as these data sources autonomy, the sharing abilities, modes and contents of the sources may change at any time. Therefore, the tight coupling solution such as federal database and data warehouse, has been unable to meet today's data integration needs. So a heterogeneous data integration platform to support these demands has been developed which can be used to complete the distributed, heterogeneous, autonomy under the data integration work.This paper takes the data centers of Dalian department of Transportation as research background. The project is designed to use a network of party and government of Dalian, it will integrate Department of Transportation under nine offices of the existing database management system, to provide a unified interface to the data query, and to implementation data integration, data sharing, and real-time data query.In this paper, at first data the theoretical foundation of data integration, and the main solution method is reviewed. Then data integration platform for the existing problems is analysed, with building the Dalian Department of Transportation traffic data center, a data integration platform architecture and implementation based on virtual data center be given. The platform uses Mediator/Wrapper as framework, uses Web services packaging business logic, uses virtual data center as the kernel, uses XML as public data integration model, and combines Hibernate with Castor to establish a conversion model between relational database and XML, solves the heterogeneous systems, heterogeneous Grammar and multiple heterogeneous data sources to XML mapping in the integrated data platform.Query processing and query optimization are distributed and heterogeneous data integration platform of key issues. It is directly related to the correctness of data and platform's availability. Aiming at the characteristics of the traffic database, based on query decomposition tree design, a query decomposition algorithm is given. This paper is based on the costs of local data sources and inter-site communications, so a query optimization algorithm based on costs is given. As the network is used for transmission of XML data models, uses the Apache Axis to create data transmission, the strategies and methods for data transmission is given. The entire transmit process can be transmit data transmit without limit with platform, language and network protocol.