Review: Service-Oriented Architecture (SOA) Compass


SOA_Compass Service-Oriented Architecture (SOA) Compass
Business Value, Planning, and Enterprise Roadmap

IBM Press (Various authors)2006

The time came for me to read-up on SOA in order to further develop my concepts of how Enterprise Content Management should fit within the Enterprise. So I starting looking for some books on the topic. A large majority of the recommended books were hundreds of pages long, not exactly easy reading for the Metro. I researched and picked SOA Compass from IBM Press.

The book is written by a collection of authors pulling on the wealth of information on the developerWorks website and from practical experience in the real world. The experience shows, as does the developer background. I quickly realized that this was not going to be easy reading and I was quite glad that I had a solid developer foundation to pull upon.

Throughout the course of the book, the authors expound on the various characteristics of a SOA environment. However it was lacking any uniformity in the examples given. In addition, it was heavy on mentioning IBM products, such as Tivoli. At times I almost felt I was reading a marketing brochure. Perhaps I was. I think that is the major flaw in this book. Listing the features of a product that solves a problem does not truly explain how to address the problem.

I did gain a new appreciation for experts in SOA. The sheer number of Web Services standards became daunting after a while. I don’t think that WS-* accurately conveys the number. I can even see why some people strongly dislike WS-*. There is a lot to it and that is even before you add any standard definitions on top of it.

Now for the positives, and there are quite a few. They did manage to cover quite a bit of ground and I kept seeing tidbits that I had read elsewhere over the past few months. I also discovered why many of those other books were so long. Once you realize that a simple list of characteristics isn’t going to cut it, you are left with either generating 600+ pages of text (as opposed to 219, including the very nice glossary), or writing sections that leave you exposed to several terms, but grasping at times for more detail.

Ok, so maybe that didn’t come out as positive as I planned. Overall, it was a good book. It was quite informative and provides frequent links to locations to learn those missing details. Most of those links are on the IBM website, but that is understandable given the Authors all work for IBM.

The best part of the book was Chapter 10 when they introduced some Case Studies. Up to this point, Web Services was presented as the end-all, be-all for SOA. That hadn’t been stated, but that was the implication. The first case eliminated that myth. The real life examples helped highlight the goals of a SOA. If they had taken one of the case studies and discussed in each section how the different SOA aspects were addressed in that solution, I think the book would have been dramatically better and more enlightening.

I did accomplish my goal from reading the book. I have a broader picture of SOA and how best to try and fit ECM into the bigger picture of the complete Enterprise Architecture. Once I get a chance to compose my thoughts in more detail, I think I am ready to start making the case for a new, SOA-based, ECM standard.