Comments on: Tutorial and pseudo-algorithm for CMDBF Query operation

By: William Vambenepe’s blog » Blog Archive » Illustrative algorithm for CMDBf 1.0 Query operation

Wed, 07 Nov 2007 05:46:29 +0000

[...] I posted an algorithm for the server side implementation of a CMDBf Query call for version 0.95 of the specification, the interoperability testing session based on that version was over and I was [...]

By: Van

Van — Mon, 13 Aug 2007 17:00:30 +0000

Some questions are answered in my latest post on talk.bmc.com, giving an example similar to your question.

On the OO question, we have avoided defining a data model, so there is no inheritance currently defined in the spec. Asset could be a record type that is associated with many items that are associated with many other record types.

By: Tim Coote

Tim Coote — Thu, 09 Aug 2007 14:51:01 +0000

When I first read the cmdbf spec, I thought “great a graph based query mechanism”. I had in mind the types of graph operators like “show me all nodes that are dominant to node X” or “get me all disconnected subgraphs of graph X”. However, this example reads more like CODASYL encoded in XML. Have I missed something?

So, I read the spec. and found a few other queries:
- What are the semantics when a sub service is unavailable - are there error values if part of the federated cmdb is unavailable and how do I define permissable conditions?

- I coudn’t find much about time. What does it mean if the values of an attribute on a CI differ, but the timestamps for the validity of the values differ (however, I’d establish that).

- What about relationships with a cardinality > 2 (eg user x is using product y under licence z)?

- Another thing that’s got me wondering about this problem space is how I idenfify, say all assets without an owner. If I had an OO model, Asset could be a superclass, or maybe I have asset as a tag on all objects that are assets. This seems like quite an easy problem at first, but the real challenge is in being able to make it work as the underlying model evolves: you need very simple initial queries and rules for schema migration.

Any help gratefully appreciated.

Tim

By: dougmcclure.net » CMDBf Draft Specification Released, First Tutorial Provides Insight

Sat, 04 Aug 2007 15:03:03 +0000

[...] William also provides some practical insight into the specification’s application in a posting called “Tutorial and pseudo-algorithm for CMDBF Query operation”. [...]