Start a Discussion
SOA
user-pic

What Do You Believe is the Biggest Inhibitor Today to SOA Adoption?

Vote 0 Votes
This question comes from Kelly Emo at HP who, very simply, asks: What do you believe is the biggest inhibitor today to SOA adoption?

11 Replies

| Add a Reply
  • Impressions like these:

    • SOA is big.
    • SOA is complicated.
    • SOA is all encompassing.
    • SOA is expensive.
    • SOA takes a long time.
    • SOA is dead.

    For all it's success, SOA has developed a bit of an image problem lately. The truth is, it doesn't have to be any of these. SOA can--and should--start small and grow organically. And if you read beyond the sensationalist headline of the last bullet, you'll find this is exactly what this was really about.

  • I'll keep it short and sweet. Non-technical manager that control the purse-strings that want to focus on tactical issues of the day and don't fully understand the benefits of strategically focused efforts.

  • Unrealistic expectations and hype, combined with lack of skills and experience, are the greatest inhibiting factors. Companies started thinking that Web Services would provide business agility, increase reuse, and reduce costs.

    I think the biggest single factor for SOA success is governance, which includes roles, responsibilities, policies, metrics, monitoring and managing throughout the life cycle. When companies start small and grow organically, governance is usually not considered until there is already a mess, and investments fail to deliver value.

    SOA success is not about the technology - it is about what you do with it. SOA is an elegant design concept. But because it is distributed, and made up of many smaller components, it requires comprehensive governance in order to deliver real business value.

  • Biggest inhibitor to SOA today? Enterprise IT teams are finally starting to get the idea of better architecture, less emphasis on the specific technology and putting governance in place, and here comes a new problem: Dependency.

    Yes, you can do everything else in SOA right, get your teams collaborating on faster iterations of loosely coupled services - only to find that your software lifecycle is getting constrained because the live services, mainframes and data these teams need to develop and test against are simply incomplete or unavailable.

    Each team, building each component that contributes to an interconnected SOA effort will experience these kinds of dependencies, in attempting to deliver version of their own functionality that meets expected service levels. To get agility we need to move teams into more parallel delivery lifecycles. Much of this has to do with more modularity in architecture and project planning, but that doesn't solve the problem alone. Virtualization is being pushed forward as one technology that can help overcome dependency.

  • The biggest inhibitor to SOA in the market today? I think there are two major ones, one of which has already been alluded to in this thread - image. SOA needs a good PR agent. The litany of "too big, too complicated, too expensive" and "oh yes by the way it's dead anyway" is deadly.

    I think the other inhibitor is the simple fact that as an industry we have not done a particularly good job of documenting SOA benefits and successes in a fashion that the line of business manager within an enterprise can understand. We default to technical discussions entirely too often and the line of business just turns off.

  • I see two potential causes for the lack of SOA adoption, the first is people forget that SOA is about "Architecture". There is so much focus on tools, specifically production integration tools, and not enough on developing the architecture that will work for the company. Secondly, and most importantly it is about Governance, Beth makes a great point above which I see all the time when talking to companies working on SOA projects, "governance is usually not considered until there is already a mess, and investments fail to deliver value". I think the market needs to come to an understanding that governance (people/process/policies) is a MUST and should really be started early in the service development lifecycle rather than when a service is put into production. If we start to see governance applied early in the SDLC, it will increase the success stories which in turn will promote greater adoption.

  • The inhibitors to an SOA implementation are many and just as varied as there are benefits to an SOA implementation. And while some inhibitors may be more universal than others, which inhibitors affect a given organization is often dependent on the organization itself.

    Here are six key inhibitors that I have experienced over the years:

    1. Unrealistic or misunderstood goals and expections— As with any substantial project, level-setting expections and gaining a common understanding of the goals/benefits amongst the stakeholders is key to the success of an SOA implementation.
    2. Tactical instead of Strategic Approach— SOA is best implemented top-down with a strategic view of the business processes and functions to be exposed as services. Bottom-up implementations too often tend to be overly focused on technoology rather than the business
    3. Not putting your money where your mouth is— Let's be realistic: The best intentions won't go too far without money being spent. SOA requires an investment in your resources (people, tools, and technology) to be successful. There is no free lunch in life.
    4. Poor Governance— Poor or complete lack of governance can lead to service proliferation, which can ultimately lead to more severe problems with data integrity, meeting SLAs, and privacy/security breaches.
    5. Organizational Power Struggles— In an SOA the focus moves away from individual applications and data stores to more business-oriented services that may cross business unit and organizational boundaries. This breakdown in application and data barriers may cause heartburn to the owners of those applications and data, who may see this as a challenge to their power see advantage in hindering the SOA implementation.
    6. Fear and Ignorance— SOA is surrounded by many misconceptions regarding exactly what it is and what it takes. Too much has been said about all the technical details/variations leading to the misguided belief that any SOA implementation must have "all that" technology and complexity. Ignorance breeds fear, which then becomes a major inhibitor to the success of the SOA implementation.

    But these are just six of the inhibitors that I have seen; there are many more. Remember, each organization is unique in its own way and the challenges it will face as it implements an SOA will depend on its unique culture and history.

  • The issue to my mind is that organizations see it all as a set of powerpoints they have seen before. In other words, as other respondents have said, it's more hype to sell product. All the boxes sit wonderfully well together in the powerpoint presentation, as they have done when the industry talked Client / Server, Object Orientation, Corba, Utility Computing etc.

    My view is that we have seen these boxes before in many different ways and as I have alluded to before, there is nothing different there until we can show _how_ these boxes talk to each other.

    We have found people quite unimpressed, uninterested and even totally negative with SOA or any of the interoperability standards. However, when we start by showing a) how interoperability standards can make the boxes talk to each other with real demos and b) illustrate the benefits of creating a SOA based on these interoperable boxes, they get interested.

    In summary, most organizations I have talked to at the operational level don't care if it's called 'SOA', 'ABC' or '123'; or whether they are using standard a, b or c; they are interested in the real benefit something brings to them. They don't believe hype anymore, they want to suck it and see to determine if it will really help them.

  • Many companies may think they have "SOA," but more likely have JBOWS architecture, or "Just a Bunch of Web Services." They may have been led to believe by their vendors that their latest solution gets "SOA" up and running for them, where actually it is just one piece of the evolution. By the way, there's nothing wrong with JBOWS, it too is part of the evolution on the journey to SOA. The fact is, few companies have actually reached the stage where they have a full-functioning SOA.It's an incremental process.

  • I agree with JP Morgenthal on this. Business and IT managers are the major inhibitors because they are afraid of SOA, which highlights efficiency (or its absence) of their work. Since service orientation moves top-down, mistakes made at the top becomes quite visible at the 'down'

    The more a work is tactical (i.e. non service oriented because tactical approach is fine with patching only), the easier to run away from made mistakes w/o responsibility.

    Service orientation is one of the best modern instruments for efficiency implementation but to reach such quality it has to be started in the Business. Until Business would concentrate on HOW it works instead of on WHAT for and WHY, the perspective of SOA are uncertain.

  • The biggest inhibitor to SOA in the market today? I think there are two major ones, one of which has already been alluded to in this thread - image. SOA needs a good PR agent. The litany of "too big, too complicated, too expensive" and "oh yes by the way it's dead anyway" is deadly.

    Thanks

Add a Reply

Recently Commented On

Monthly Archives

Blogs

ADVERTISEMENT