The discussions and news around SOA are clearly moving on from cool features towards tackling the nuts-and-bolts issues of how to get it done as can be seen from a few of the items that have appeared in the last week or so.
Joe McKendrick does an excellent job of pulling together a number of quite contradictory surveys attempting to measure exactly how great SOA adoption is today – with IDC coming in at the upper end with a survey claim that 90% of respondents were on the way to SOA.
His conclusion is that most organizations claiming to have deployed SOA are only in the early stages:
These statistics vary wildly, and lead one to conclude that SOA must be in the eye of the beholder, because one company's "SOA" may be, in someone else's view, a Spaghetti-Oriented Architecture or "JBOWS" (Just a bunch of Web services).
What all of the statistics show is that there is huge awareness of SOA and most people feel it is something that they should be doing. Clearly, there is absolutely no way that 90% of organizations or anything like it are already SOA based.
Accenture were recently reported as committing $450m to a new SOA technology lab in Chicago.
Meanwhile, Don Rippert, CTO of Accenture was in Hong Kong and outlined his view of the 4 phases of SOA development:
Phase one: "Organize and Strategize": get management buy-in and plan for SOA transformation
Phase two: "tactical implementations": embarkation on SOA projects and the conversion of applications into web-services, and the creation of business processes from such.
Phase Three: an emphasis on strategic and business services; consolidation of processes and services in creating an Enterprise Service Bus.
Phase Four: "SOA is industrialized," he said. This will involve cross enterprise processes, federation, predictive IT and business insight in real time.
Focused as you would expect on the high level, I am not sure there is much to argue with these statements. Each phase has its own challenges – in particular moving from phase 2 to phase 3 and beyond. That is where organizational and role changes need to begin. For instance, Wachovia took a product management based approach to ensure that SOA remains responsive to the needs of business.
And finally, David Linthicum has tackled the sensitive issue of whether the EAI dinosaurs can really dance to the SOA tune:
Many are finding that the more traditional *EAI technology is coming up short when you consider the new complexities of SOA.
And
Perhaps the core issue is innovation. While a commodity in the newer venture funding startups that don’t need to declare an EPS, are just cost to the larger companies, and thus to be avoided.
As a person who spent almost 5 years in a SOA start-up (PolarLake) I clearly believed and still believe that there is an opportunity for well-funded SOA start-ups to take on and beat the established vendors. However, some of the established vendors are already ‘evolving’ fast and their SOA offerings are already as impressive as the best of the start-ups. As so often happens in a new market, expect to see a couple of the familiar names disappear (as has already happened for instance to Mercator an old-reliable of the EAI marketplace) and a slot or two emerge for a new player.
Subscribe to feed
Leave a comment