Joe McKendrick recently wondered if there is too much "whining" and not enough action on SOA? in a ZDNet post. He was responding to post by Nicholas Petreley of CIO.com on Less SOA QQ and More SOA Pew Pew.
Let me translate: QQ means whining and Pew Pew means action. To be more specific the urban dictionary offers some definitions of Pew Pew: Sound made by lasers, usually related to star wars. And also a common term in the world of Warcraft most typically meaning to do damage, or perform a combat action.
So is Nicholas telling us to ready-fire-aim? Joe points out that there is already a lot of SOA action. He adds, "there is nothing wrong with a good rousing debate on what SOA means for an organization. Perhaps there should have been more debates like this before companies invested millions in ERP and CRM systems in the past." So if whining becomes debating, this can be a good thing.
SOA can create agility and some incredible efficiencies, but there are a lot of issues to resolve when plunging in, unless you want to do a lot of pew pew (e.g., damage). For example, because of the complexity that comes with orchestrating multiple services, work out governance issues in advance.
Still, the QQ (whiners) about SOA may be surprised to find that they have development and integration teams that are already soaking in the SOA pool - creating services-based architecures - even if the IT management isn't classifying the efforts as such. We have seen this phenomenon at many large companies.
We also suggest looking at three key issues to consider before attacking an SOA strategy.
- First, acknowledge the fact that this is a heterogeneous layering of technology, so plan to verify outcomes in other layers and systems of record beyond the one you are building and testing.
- Second, enable more parallel activities to leverage assets more rapidly and take advantage of SOA. This is where Service-Oriented Virtualization comes into play -- if you can prune off those interdependencies upon other technologies by Virtualizing the rest of your architecture, it enables each team to be productive throughout the lifecycle.
- Third, we cant test as an event along a life cycle anymore, and think that this is enough. You need ongoing Continuous Validation, because services are coming from multiple sources which might change. Without this validation capability, it is impossible to know if your SOA Policies are actually being carried out in the real world, as there are multiple changing elements that can affect the workflow.
The factors can help with the SOA debate -- which is a better approach than either QQ or pew pew.
Comments