Stephen A. Fuqua (SAF) is a Bahá'í, software developer, and conservation and interfaith advocate in the DFW area of Texas.

Evaluating SOA for Enterprise Architecture

September 10, 2008

A few weeks ago I started trying to better understand the concepts behind service oriented architecture (SOA), how I can apply them to my work, and if doing so is actually a good idea. When I first started hearing about SOAP years ago, I understood it as an alternative to objects — you build and consume services instead of objects. This sounds great for highly distributed environments, but also sounds like a performance nightmare in a business environment that does not need distribution. So I largely ignored it, and assumed I could ignore SOA as well.

In the last few years, it has become more clear that SOA also allows the application designer to open business logic to several different platforms. Essentially these two statements are the same, but it seems that they can easily lead people in different directions – the former eschews objects altogether, while the latter statement is agnostic to the manner in which the business logic is built.

In the process of re-evaluating the system architecture I have developed at work, I can see potential for taking advantage of the 1-to-many relationship of business logic to user interface platform. So I’ve been slowly, gradually trying to evaluate SOA's usefulness in my still evolving business logic architecture. As a .Net developer, this slow process has included trying to better understand the role of Windows Communication Foundation (WCF). Looking for some thoughts and patterns for integrating a service-oriented approach with an object-oriented one — which seems imminently reasonable when treating the service as a Remote Façade — I found some interesting discussion diving into these issues:

  • What is Service Oriented Architecture, a very early article on the subject (2003), published by O'Reilly, definitely leaves one with the impression that the two concepts are incompatible.
  • In Pro WCF (Apress, 2007), the authors make it sound like SOA is an alternative to OO, with statements like “Until now you probably have been creating applications utilizing an object-oriented programming model. Service-oriented architecture (SOA) is a fundamental shift to dealing with the difficulties of building distributed systems.” While this statement alone could fit with either way of looking at SOA, the book (of which I’ve only skimmed a small portion) continues in several ways to make it sound like an adopt-one-or-the-other proposition.
  • Martin Fowler seems a bit dismissive of SOA, and in his Patterns of Enterprise Application Architecture (Addison Wesley, 2003), he makes no mention of SOA. He does use one example of web services as a façade.
  • David Ing does a nice job of classifying the two SOA approaches above – plus others – and giving them names. So far, I am in agreement with him that its value seems to be in “Services as Integration Gateways”.
  • For now, I'll end with a pithy set of comments on the Creative Karma blog, which has the most digestable explanation I've found for what is actually fundamentally different between the two architectures: encapsulation.

In the end, I still haven’t determined if it makes sense to start re-engineering my applications for WCF, though I have been convinced that there is no real point to talking about "SOA vs. OO" (yes, I realize I'm  very late to this party, it is so 2004). I suspect that will have to be a corporate decision — are the tradeoffs worth it? But in the meantime, I feel a bit more knowledgeable about the domain, and better prepared for making a particular case in the light of actual business requirements.

No TrackBacks

TrackBack URL:

Leave a comment