Enterprise Software

Five fearless predictions for SOA 2009

ZDNet blogger Joe McKendrick offers five predictions about the shape of things to come for service oriented architecture in 2009.

 

This is a guest post from Joe McKendrick of TechRepublic's sister site ZDNet. You can follow Joe on his ZDNet blog Service Oriented, or subscribe to the RSS feed.

The year 2008 was the best of times and the worst of times for service oriented architecture. SOA seemed to become more and more pervasive across companies, and many companies in general seemed positive about the results being delivered so far. At the same time, there has been no shortage of anti-SOA backlash and skepticism about the return on investment and business value of SOA.

What lies in the year ahead? Here are some fearless predictions about the shape of things to come:

1) Economic turmoil will return SOA to its roots - bottom up, incremental. Uncertainty in the economy will carry through 2009, but we'll move into the recovery stage. However, this won't change the razor-edge competitive environment - companies will continue to seek solutions that help streamline and cut costs. This is a natural role for SOA-based practices -- remember, Web services and SOA were forged in the wake of the downturn of 2001 as a way to increase IT efficiency and value to the business with minimal additional investments. There will be fewer big-bang SOA projects rolled across the whole enterprise, and many more incremental, bottom-up efforts -- many of which may be under the radar. More guerrilla SOA, if you will. SOA will also make it easier for companies in mergers, acquisitions, or restructurings to reconstitute themselves. 2) Vendors will de-emphasize SOA as a distinct "product" offering. There will be less hype about SOA, but that doesn't mean it will have gone away. New solutions and applications will have service-oriented aspects. Cloud offerings will be built in accordance with SOA principles. There won't be a lot of start-up vendors pitching SOA solutions, but plenty of start-ups will be offering Web 2.0-type and cloud-based services, which will be underpinned by SOA principles. 3) Internal clouds and micro-outsourcing. There has been a lot of industry discussion about the "internal cloud," in which services are provided to users and systems within organizations. The beauty of internal clouds, of course, is that they offer more control over applications and data. Clearly a natural role for SOA, which will be the backbone of any emerging internal clouds. As part of this role, expect to see SOA play a greater role in grid computing and virtualization as well. Externally as well, more SOA initiatives will include services from outside the firewall -- a sort of "micro-outsourcing" of application functionality. 4) More attention to the data element. Companies may do a great job of streamlining and leveraging processes, but often ignore the quality and viability of the data flowing through these processes. Which makes for unsuccessful processes and unsuccessful SOA. At a time when "competing on analytics" will make a difference, companies will be paying more attention to the data their SOAs are serving up. 5) More Web 2.0 tools in the SOA world - and new governance issues. The convergence of Web 2.0 and SOA practices means more interesting approaches to old problems, such as gathering business intelligence. And mashups - many of which may even be designed by users themselves under the watchful eyes of IT -- will become the default composite application of choice for accessing services both internally from SOA-enabled systems and externally. Organizations that have already been wrestling with governance issues for SOA will find themselves with the question of how deeply to regulate mashups and other Web 2.0-ish activities.
3 comments
Justin James
Justin James

... is that it will continue to shrink as a class of product, and simply become an adjective that some products have as a feature. It's like "open source" 5 - 10 years ago. As it matures and the hype cycle winds down, and people can see what's reality and what's spin, they'll see it as not the sole differentiator of a product or service, but a feature. And thank goodness when that process get's completed. :) J.Ja

Dr Dij
Dr Dij

reading a blog on an SOA aite, the discussion went something like this: companies are missing out the savings, flexibility, other advantages of SOA they are stupid not to implement it they are losing competitive advantage, blah blah synergy blah blah.. so there still is hype I was thinking to myself: companies KNOW instinctively that any system has costs, and complications. the newer it is the more the costs and complications. What these bloggers were not saying (possibly not thinking) is that below a certain level that overhead is HUGE! The backbone of the economy is small-ish companies of all sizes. Does Joe's Pizza Parlor with one store need SOA? Does Mike's string of 10 need it? DOes it need SOA when it reaches 100 stores or $xx millions? Companies themselves are the best judges of what money they have left to spend on techniques / technology and if it is relevant to their biz. SOA is useful particulary if you write or enhance alot of your own SW. But many companies are trying to reduce this to near zero or at least only addons or reporting from canned systems. Does it need SOA if it buys a canned ERP? If you DO need it then there are large-ish people costs - training pgmrs to rethink development, SOA governance to be sure is done right, lots of new security ramifications, interfacing SOA based systems with vendor / client / legacy / other systems, .... and reqt for people to think this kind of stuff thru, not cheap either. or endless BIG # computer consulting firms to ply you with ideas and someone with your company's interests to make sure they don't sell you a boilerplate solution that doesn't fit. So glad hype will tone down. Why should a company even NEED SOA as a separate 'THING' if they are not buying or selling SOA. Instead they may just need their company's base systems to be 'SOA' based. I'm saying that this is like having a brand new engine that everyone is raving about, and then wondering why your company doesn't have this 'engine'. Chances are your company doesn't use 'engines' directly for anything. They buy delivery trucks that have whatever engine the vendor sells it with, nice if it is better or more efficient or easily upgraded. They use engines or motors in production equipment and elevators but never would they buy an engine for itself, and this is what the hype sounds like to me.

Justin James
Justin James

Quite right about the relationship between SOA and the size of an organization... SOA by its very nature is an architecture-heavy beast, which in turn allows the applications to be quite lightweight. Many, if not most organizations are simply unable to realistically put together quality SOA that will be used enough to justify the efforts. But try to tell that to the SOA hypesters! J.Ja

Editor's Picks