IBM eyes services-based role for J2EE

By the first half of this year IBM Corp. will aggressively extend the Java Enterprise Edition 2.0-based architecture of WebSphere to a services-based architecture that allows developers to build applications with integrated workflow, business rules, and network-based transaction capabilities.

Although IBM does not plan to de-emphasize J2EE in any way, company officials on Monday said they believe that J2EE-based technologies alone are not sufficient to accommodate the emerging on-demand computing environment that is central to many of its strategies going forward.

"In the on-demand applications world, J2EE alone is like having a heart without the lungs. Putting a Web services veneer on top of J2EE is just not sufficient, according to Scott Hebner, IBM's director of marketing for WebSphere, in Somers, N.Y.

Along with the integrated workflow and business rules capabilities, IBM will also deliver a more complete Web services stack in the upcoming version of WebSphere 5.X, Hebner said. The improved Web services stack will include the Web Services Invocation Framework that allows for network-independent interaction among Web services, a precursor technology to a business process execution language, and the ability to have compensation patterns that allow users to back out of any given Web services workflow.

Hebner said the applications built using the new architecture will inherently be able to integrate business processes across an organization dynamically so users can treat them as individual business services. This is something that cannot be accomplished using just J2EE and Web services technologies.

"You will be able to take these applications and choreograph them into network-based workflows that have transactional integrity without having to understand the underlying complexity of how they were built or where they sit on the network," Hebner said.

The services-oriented architecture of WebSphere is only one of three major pieces that will make up what IBM officials believe will be the Next Generation Application Server. The other two pieces are sets of autonomic capabilities such as self-healing, self-managing, and grid computing capabilities.

"In the next generation of application servers, there needs to be more advanced clustering and failover capabilities, and those will all evolve around grid standards," Hebner said.

Explaining the advantages the next generation of application servers can bring, Hebner said developers will be able to look at a palette of business services whether they are written in Cobol, EJB, or Windows and take them to compose and choreograph them into entirely new applications.

"You will see us being very aggressive around this idea over the course of 2003," Hebner said.

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

More about EvolveIBM Australia

Show Comments