After a very long and very cold winter, I’m counting the days to Easter break, which I will spend sitting comfortably on a warm, sunny beach. The last time I saw the bats was on the beach, and this time I’m afraid they’ll be more frantic than ever.
You see, I’ve been thinking a lot about CMDBf and workflow vs. my favorite subject, Service Monitoring Intelligence. It’s a confusing picture, which perhaps is why those nasty buggers are flying around again…
The CMDBf is a standard released by the Distributed Management Task Force (DMTF), which “simplifies the process of managing related configuration data stored in multiple CMDBs and MDRs and supports the creation of an ITIL v3 Configuration Management System (CMS).”
So are we any closer to Nirvana? Well, I don’t think so. In fact, I’m a bit fed up with this federation crap.
I can’t explain the technical details of CMDBf, but what’s most unfortunate is I don’t think they matter. If your primary supplier is an 800 pound gorilla and doesn’t want to federate their CMDB data, then I suspect they can make it pretty tough for you to do so (unless you are also an 800 pound gorilla).
But what’s got the hair on the back of my neck sticking up again is that on the one hand, we want the ability to tightly control our processes, and that includes being able to track things like Incident, Problems, and Changes as they relate to Configuration Items, but on the other hand we need to make sense of Events, which is more about real time information and correlation.
So ‘federation’ (a la CMDBf) would seem to be the answer to our prayers — the best of both worlds.
Then the reality hits. Those big gorillas are not about to give up (what should have been their focus all along) on event management. You’ve got all your workflow tied up in their ‘CMDB’, and they’ll be damned if they’ll let someone else crash their party… a food fight if ever there was one.
So customers are faced with a choice of getting sexy integration between existing workflow systems and all their Incidents Changes and generally poor real-time event information (and even less event intelligence) … or a good event management product that is less ‘integrated’ than the CMDBf spin we hear from these guys.
It seems like such a cruel irony. If we’re able to manage Events really well, how much does that help us mange workflow? A lot, I think. How many times have I heard that “most Incidents are caused by Changes” or “it takes 8 minutes to fix an Incident but 8 hours to diagnose it”?
Even ITIL states that the Event Management process is “the basis for operational monitoring and control, service assurance, service reporting and service improvement.” So what the hell are we doing focusing on workflow? Why has it taken the big gorillas five years of CMDB and ticketing madness to now come up with the brainstorm that we need ‘run-time’ service models and ‘real-time’ CMDBs?
Customers may want to take another look at what their process maturity and tool integration requirements really are before paying too much attention to this CMDBf spin. One the one hand, it’s a promise of multivendor integration. On the other hand, you’ve heard this before… whose interests are you really hearing about, yours, or someone else’s?
You know what I think.
There’s a need for decent integration between your event management tool and your workflow tools, certainly generating Incidents. It might even be really neat to achieve the vision of CMDBf. But the reality is most of us are headed for a cloud-based, multi-service provider nightmare that is going to require real intelligence and transparency across service providers.
Having all our tickets neatly in sync and no idea what’s going on in all those clouds doesn’t sound like much fun to me.
Getting stuck in more of these visions of Nirvana just doesn’t seem like the way to go. Federation won’t get you very far if you have to give up the level of event management intelligence that you need. In fact, I thought that was the whole point! I get to pick exactly what I need and just ‘federate it’!
If you mange events well, workflow gets a lot easier. In fact, I think high enough levels of event management intelligence may go a lot farther to help strained ITIL processes than ‘federation’, with a hell of a lot less effort.
If you want a ‘simple on-ramp’, I’m not sure CMDBf is really so simple, and, perhaps more importantly, I’m not sure your vendors will really go out of their way to help you integrate someone else’s product.
Even the CMDBf states that “the amount of data to federate depends on the services to be supported and the level of configuration control desired. … The only guideline for what data should be included is business relevance.” Balance the talk about ‘federation’ with why you want to federate in the first place; to give you the flexibility to add other tools to your CMS.
The speed and complexity of today’s virtual technologies and cloud-based operating models demand greater services orientation, real-time management of end-to-end IT Service dependencies and an improvement approach that can compliment existing process-oriented improvement programs.
So the next time you hear one of those big gorillas talk about federation, show them what you selected and tell them to… federate this!