When I started at my current place of employment, we were just transitioning support of a medium sized UNIX environment from another company which had been supporting the installation previously.  Unfortunately there was no system automation in place and all changes and updates had to be done manually.

Having been a UNIX systems engineer for a long time, I'm not a fan of manual configuration.  It's time intensive and error prone.  So I set about configuring a cfengine installation using Subversion for revision control of the configuration files.  My team mates were not familiar with cfengine or Subversion, so I documented their use on our internal wiki.

I've had occaision to set up fresh cfengine configurations at several contracting clients now, and it occurs to me that while cfengine is incredibly powerful and useful it can be a bit overwhelming for new users even if they are experienced UNIX administrators.

In my copious free time (yeah, right!) I should start a consulting business around system process automation using cfengine with revision control for more managable and scalable infrastructures.