Note Joe doesn’t ask if it is required, necessary, or useful – just affordable.
It’s an interesting question. Here are some additional thoughts:
1. EDA and CEP does not replace BPM, SOA or databases.
Events still drive processes (human) and services (IT) – and indeed CEP can be thought of as an event driven process or service. And events end up being needed to be stored for historical analysis / analytics – so a database is still necessary somewhere in the architecture…
2. Are EDA and CEP less affordable than SOA and databases?
Not really. The technology is getting pretty mature now – consider the ubiquity of messaging middleware – and CEP incorporates the “best practices” of IT from SOA and database worlds – e.g. declarative rules, model-driven, object-oriented, distibuted storage, event streams and patterns … but this is still a “value” metric, not an “affordability” one. The affordability comes down to development and deployment costs versus ROI: development can be quicker, and deployment can often avoid a boatload of application server nonsense (i.e. administrative stuff that is not business logic). And the ROI of the ensuing “operational intelligence” can of course be immense.
3. Is “open source” the saviour of affordability?
I saw this was one (and a typical) response to Joe. Again, not really – open source tooling has its place, especially for educating ourselves. But affordability is a lifecycle issue not a development tool cost – and I am reminded by the customer who spend a few months trying to build something with an open source tool that was solved in a few weeks with TIBCO BusinessEvents. And most CEP vendors provide evaluation copies that mitigate up-front costs. Remember “open source” is just one more business model option for “vendors” (where the “vending” is of support, maintenance, services etc).
4. Is there any proof of affordability vs value etc?
I can’t say I’ve come across many folk who have said “nice but we can’t afford it” – from a cost perspective anyway. Often IT budgets are consumed in getting existing IT systems functioning – the affordability is affected by the non-affordability of existing IT infrastructure. Most EDA and CEP systems of course integrate rather than replace conventional IT, and there is an additional emphasis – not burden – on IT architects in organisations to understand what fits where. But architects I have met like EDA and CEP (albeit I am unlikely to meet uninterested architects!). Some use cases show pretty powerful ROIs.
5. If EDA and CEP is affordable, what is the problem?
Entrenched views. CIOs bought up on data-first mentalities. Not enough “thinking outside the box”. Anything that resists paradigm shifts. Which is fully understandable, and why there will be steady, not revolutionary, growth in CEP and EDA markets. Although the published growth data looks pretty impressive!
The full discussion is on ebizQ here.
