RSSExpert analysis, advice and prognostications about Service Oriented Architecture and distributed computing.
Our bloggers: Mike Kavis is a veteran Chief Architect with over 23 years of IT experience including distributed computing, SOA, BPM, data warehouse, business intelligence, and enterprise architecture. Former applications developers Rich Levin has been implementing, advising on, and writing about information technology for over 20 years, covered computer technology for CBS Radio and hosts the popular "PC Talk" show. Nicholas Petreley is a former programmer and consultant, has worked for InfoWorld, Computerworld, LinuxWorld and Network Computing World, webzines, and serves as contributing editor for CIO, focusing on SOA as a primary area of coverage.
'A' is for Architecture
Keywords: SOA, Architecture
There is no shortage of industry research that validates momentum around SOA. Likewise, there is no shortage of books, papers, pundits and bloggers who espouse or otherwise agree with one or another definition of SOA. But the fact remains that any discussion of SOA quickly devolves into an IT version of David Ben Gurion's political maxim: "Wherever there are two Jews, there will be three opinions."
Likewise, wherever there are two IT professionals, there will be three opinions of SOA. It's odd, because at the end of the day, SOA is a simple concept that anyone in IT should grasp in a nanosecond. It's also troubling, because unless we agree on (read: universally understand) the definition of SOA, the research and conclusions extrapolated thereof are simply garbage out, because the premise was garbage in.
For example, last week I cited an AMR Research report that projected IT investment in SOA will hit $52 billion over the next four years, with 77 percent of IT shops building SOAs. But if some of those surveyed believe SOA is a floor polish, while others believe it's a dessert topping, the projection is flawed and, as such, not actionable.
I'm not singling out AMR. But having interviewed many a CIO who is leading the charge on a SOA implementation (several of whom I'll feature here in podcasts over the coming weeks), I'm willing to bet that the folks AMR interviewed for their survey don't agree on the definition of SOA and, further, that many are operating under a flawed definition of SOA.
It's not a stretch to say this cognitive dissonance affects much, if not all, of the punditry and prediction bandied about on the topic today. It's certainly shaping the response of IT technology vendors, who are chasing, delivering and evangelizing platforms designed around, in some case, wildly divergent approaches to SOA.
For global IT to truly benefit from SOA, we must first understand what SOA is. Let's start with the 'A' in SOA, which stands for architecture; specifically, in the case of SOA, a software architecture. Rather than spark a debate about what constitutes an architecture, I'll lean on the definition adopted by the IT Infrastructure Library (ITIL).
