Eric Roch <http://it.toolbox.com/people/eroch/> (Chief Technologist) posted 20 hours ago
http://it.toolbox.com/blogs/the-soa-blog/soa-alive-and-well-30282?subtype= As debates continue on the viability of SOA, we continue build service oriented applications and software vendors continue to service enable their offerings. Application vendors such as Oracle and SAP are aggressively adding services and industry specific process orchestrations to their SOA frameworks. New releases of technology software such as content and data management have added services and process orchestration capabilities to their software. Business process management systems (BPMS) and rules engines have enjoyed renewed interest with the addition of services support. In fact, the value proposition for BPMS has been enhanced with the capability to integration legacy systems (and new components) to the BPMS with services. Everywhere I go I also get asked about open source tools for SOA. And these tools are maturing. There is a lot of bottom up working going on with Java frameworks to build business services. This work does not get a lot of vendor attention because there is no big software spends. Certainly some of these services are technical widgets but I see successes with business services all the time. Even Web 2.0 technologies such as RIA and Mashups have the capability to include services and XML if not out of the box at least through a simple to implement framework. This gives us the ability to integrate web resources and enterprise class services within the same application, thus, persevering internal security, QoS, transactions and so forth. *Like it or not SOA is here to stay and continues to evolve.* The fact that many companies have gotten SOA wrong on the first cut is not surprising but does not change the march towards a more modern architecture. Didn't many companies get ERP implementation wrong in the early 90s and yet ERP is now the dominate application system for virtually every large company? This fact also implies that an ERP strategy must also include a SOA strategy - for example what is the state of an ERP vendor's service catalog, how will services within the ERP evolve over time and how does this impact integration, application development, migrations, module acquisitions and so on. I have to ask if the work we are doing is not SOA, then what are doing? We are building business services and the infrastructure to support them. We are improving our governance and organizational structures to support shared services. And we are moving to a well defined layered architecture with utility services the integration of the layered architecture. Our designs and plans are obviously architecture, but are they service oriented? Many of us are building discrete components for tasks, utilities and even business entities. Do you still believe in SOA? I do. Be patient the migration is going to take a long time! Enjoy reading it All the best Ashraf Galal
