Is SOA dead or not?
Well, boys and girls … inaugurate this blog as a way to express some ideas about IT and, if possible, share some knowledge. Much of the knowledge that accumulates in IT does not come with critical analysis, ie, managers, architects, developers and almost everyone involved in any way with IT not do an analysis of how important a tool / technology is important for their business / system.
And example is the technology that is mentioned in the title of post:
SOA serves for what? And to whom?
SOA is essentially an integration technology (EDI, EAI and other acronyms go hand in hand with SOA). This means that any procedure in SOA should be carefully thought out. Prepare a system to embark on SOA is a task that requires concern with several aspects:
- What other applications might need on my system
- data that will be exposed will be sufficient / excessive?
- who use my system will be inside or outside the organization? How to ensure that only the right people see the data exposed by my services?
- How to ensure that the system will always flexible
These are the concerns that surface when using an SOA implementation. And, somehow, are responsible for the failure of technology to return the ROI and agility that is promised by the vendors of SOA.
But these concerns are in development, right? How to ensure that developers will have these concerns? The answer is simple: the developers not going to worry about it. And nor should they. The key to effective implementation of the SOA Governance SOA.
Apparently, it’s simple. And it is. The problem is that some companies refuse to systematically implement a governance strategy for thinking that it is not necessary, or because you think the developers alone can meet the data requirements. Turns out, no, developers are not paid to think about the flexibility of future systems, these systems only. And so they do. Then, the deployment of SOA failure every time a project has no SOA governance.
This is a very simple logic. 100% of SOA projects that I know that they had failed SOA governance. And then, some people may say “is, but almost all cases of SOA governance have also fail.” And those people I say nothing, show evidence. IBM has a page of success stories for anyone who thinks that SOA is dead.
I say the following: SOA is not dead, nor die. Just let people know just apply the knowledge properly.