Paul discussesOracle’s comprehensive strategy and roadmap for it’s [tag]Fusion Middleware[/tag] [tag]SOA[/tag] stack. After BEA’s merger with [tag]Oracle[/tag] for $8.5 Billion,  it was predicted but one of the most awaited disclosures in the [tag]SOA[/tag] world. Here is my take on Charles Phillips’ disclosure in his 105 minute webcast:

1. There will be no forced migration for existing BEA customers.

I assume, Oracle will make it more tempting for you to move into Oracle suite rather than forcing you out of BEA. As a customer, I would never like to stay with the same version for entire lifetime. There are lots of good things happening in this world and to remain competitive, I feel its very important to be continuously upgrade to newer technologies. Oracle has beautiful plans to bring in Fusion applications. This should make things a lot easier for existing BEA customers to scale up to a pure SOA environments.

2. BEA Weblogic Server Java application server becomes Oracle’s strategic J2EE containerand has been integrated with Toplink and Coherence grid. 

This is a good news. BEA is one of the best J2EE containers in the market. Leveraging BEA App server will help Oracle boost all it’s application server offerings on a robust and proven foundation.

3. Integration of Oracle ESB with BEA’s Aqualogic Service Bus.

Implementations that I have seen, use BPEL manager for transformation and Oracle AQ for messaging. ESB has been one of the holes in existing Oracle SOA suite, which has always bothered me.  I think the Oracle has realized that too and they are trying to fill this with again a proven offering from BEA. This also means, if you have not segregated transformation and messaging logic in your implementation, there will be issues for existing customers that are using Oracle FMW. Overall this was expected and is the best way to do it and therefore they get a 4 star from me.

4. BEA’s Enterprise repository becomes Oracle’s SOA governance repository. Oracle Service Registry to play a role of UDDI compliant registry.

I may be wrong, but as per my information Oracle has been  bundling HP Systinet into the SOA stack until recently to cover SOA registry and Governance offering. Time now to replace it with it’s own. The question that still remains is how strong is BEA SOA governance product? I still see gaps in the governance capabilities and therefore I am limiting this to 3 points.

5. Oracle BPEL Process manager to continue as SOA service orchestration platform.

I don’t see a reason why it shouldn’t. Collaxa engine was one of the best when Oracle bought it and it is still one of the best pieces in Oracle SOA stack. I am happy they are sticking with it. Oracle has done an excellent job integrating it with a large number of adapters and has been pretty good in all aspects. This deserve to take all 5 points.

6. Extends Oracle Process Manager adapters to Weblogic Integration.

A clever move indeed. One of the weakest areas of BEA Weblogic has been lack of wide variety of adapters. Oracle resolves this cleverly by extending it’s rich adapter set to BEA. Good move, but does that mean BEA Weblogic integration remains the way it is? Lets wait and see…

7. OTN will merge with BEA Dev2Dev and BEA Arch2Arch. 

I appreciate Oracle for their effort to merge these communities. This will empower the communities tremendously and they will be able drive Oracle to close the integration gaps, which are inevitable before the two suites finally become one.

Challenges that I foresee in this strategy

1. No matter how much one plans for integration, its always looks easier in paper than in reality. It will make the Oracle think-tank sweat when they start thinking about integrating two product stacks. I foresee no less than 2 years of time before we see seamless integration.

2. When it comes to SOA, Oracle has been a comparatively recent entrant into the market. The challenge will still remain with Oracle, to prove their mettle with end-to-end Fusion middleware case studies and references.

3. I am expecting Oracle to provide a migration roadmap to all it’s customer who are/have already invested in Oracle FMW and BEA suites. With so many things changing, it could be a challenge. I as a customer could feel jittery when it comes to investing my money in a SOA stack that is going to radically change in coming years.

Oracle has finally taken it’s way to it’s ambitious journey to provide an interesting product suite, which could easily be the best. It still needs to fill in some holes in it’s stack in areas like SOA Quality management. Oracle needs to further strengthen its Enterprise Manager to cover SOA operations area.  It would be an interesting thing to watch Oracle is able to do all that and yet align with the promises it has already made on Fusion applications.