Service Orientated Architecture

Although typical SOA implementations hide enterprise data behind service interfaces, the integration of that data is still fundamental to any successful SOA implementation. Key data access issues:

• Consolidation of data between multiple applications
• Ownership of enterprise data by services
• Interface definitions

There are several design patterns aimed at supporting enterprise data in SOA implementations, some of which are well established and some are emerging. An example pattern is the Enterprise Service Bus or Data Bus and Transformation Manager (TM) can help to significantly reduce the cost and effort in delivering data integration components for most if not all.

The TM toolset is an enabling technology offering significant benefits in designing and generating data integration components. TM generates self contained Java components that can be deployed where most convenient within an SOA architecture. The generated code makes no special demands on the Java environment in which it runs, and has been successfully deployed in various J2SE, J2EE/JEE, Web Services and proprietary server environments.

Security, threading, connection details etc are handled by the container, and data can be processed using either a pull or push oriented API. The streaming nature of the runtime ensures performance and scalability. In addition to the many usual data sources TM supports, such as XML, CSV and database, it can directly integrate to or from an existing Java object model, or use existing Java libraries, leveraging rather than replacing existing business logic.

A good introduction to the overall use of Java within an SOA environment can be found at : http://www.onjava.com/pub/a/onjava/2005/01/26/soa-intro.html?page=2

For further information Contact us