Glenn O'Donnell over at Forrester had a great post titled “Is a CMDB even possible?” that I just came across. As you may know, CMDB stands for Configuration Management Database and is typically defined as a single repository holding all configuration items (systems, applications, etc) and their relationships. The idea of a CMDB has been a core tenet ITIL for a long time. Late last year I invested a substantial amount of time over a couple months pursuing (and achieving) the ITIL Service Manager certification (ITIL v2). In earning that credential and interacting with a lot of other ITIL trained people over the years at various customers, the CMDB concept seams to be the one concept that resonates the most with people. I think this is because it is really the only area in ITIL v2 that has a bit of a technical nature to it whereas the primary focus is process. Since a lot of folks that participate in ITIL training are IT folks, I think they tent to naturally gravitate toward the technical.
In any event, while I haven’t had time to dig into ITIL v3 in any detail yet, one of the big changes is that it moves away from evangelizing a single, monolithic CMDB and toward a Configuration Management System (CMS) that may be made up of several different management systems. Glenn’s article goes into the reasoning for this and he has some thoughts on where this might be going in terms of federating different management systems.
The rush toward the holy grail of a single CMDB consumes a lot of people and resources when the newly “indoctrinated” come back from ITIL training. I think that outcome was the biggest flaw in the definition and delivery of ITIL v2 and I’m glad it has been changed in V3 to a much more feasible approach.
In terms of the Microsoft stack, obviously System Center is where these concepts are and will be instantiated. System Center Service Manager will be bringing a lot of capability in this space. This week at MMS there are at least 9 sessions on Service Manager.