Please allow me to rant about Google Code for a minute.<p>1. It's hard to tell which projects are official Google projects and which aren't.<p>2. The total lack of real names is disturbing to me. e.g. look at <a href="http://code.google.com/u/imsnah/" rel="nofollow">http://code.google.com/u/imsnah/</a> -- it's completely useless.
I've been using this at work to rebuild our internal infrastructure for test, file serving, etc. I was intrigued by its ability to manage disk redundancy via DRDB, though I haven't really tested the instance failover features. So far, I've been focused on getting it working nicely for Ubuntu guests. It's pretty cool to bring up a fresh machine with the latest packages in a few minutes. It'll really come in handy when I start working through configuration management and deployment automation for our apps.