Ron Palmer is joining the increasing ranks of CMDB critics and skeptics. He makes a number of good points about CMDB scope creep and lack of requirements, and lack of traceability to business value.
I notice a common theme in too many CMDB writings and that is a seeming lack of exposure to the practices of enterprise architecture, in particular data architecture. Many if not most of the problems being identified by the critics can be resolved by applying the same techniques to understanding the CMDB, that we would apply for any business facing system. You need a detailed and pragmatic process architecture, data architecture, and systems architecture, with traceability between them - stuff that is hard work to do. (Of course, I think that I have completed a credible first cut in my book.)
Until we have a common reference model, we're all just talking past each other.
-Charlie
