The SDO Data Access Service Component For Relational Database

Enterprise applications are built on business data. How to access business data efficiently and effectively is the key issue of enterprise application development and integration. Traditional data integration solution based on middle-ware often uses XML as global data model. But there is a mismatch between the object model in OO world and XML schema, thus much time is spent on marshalling and unmarshalling XML data.To solve this problem, based on the research on the programming model of Service Data Object technology (SDO), this paper presents a solution that use SDO as global data model, and describes the design and implementation of the data source wrapper for relational database, which is a component of the data integration platform and can also be recognized as the Data Access Service component for relational database (RDB DAS). RDB DAS performs bi-direction data mapping between database relation model and SDO object model. It executes SQL queries, returns results as a graph of DataObjects and reflects changes made to a graph of DataObjects back to the database. It supports both dynamic and static data types. Generating SDO data types dynamically according to database schema at runtime obviates the need to define data transfer object during development time. Compiling time static data type support provides better performance and security. The cache strategy plug-in support reduces database access to improve performance. Considering the concurrency problem in a multi-user environment, RDB DAS utilizes the optimistic concurrency control mechanism to provide better scalability.In summary, RDB DAS simplifies the access to relational database in SDO programming model. It is not only an important component of the data integration platform based on SDO, but also provides a powerful tool for SOA application development which use SDO for data representation.