I know firsthand how excruciating it is to toil while the downtime clock keeps ticking. The first thing to go out of the window is the ability to think critically and examine alternatives to the most obvious approach. You should take a well-earned vacation to get it back.
From your description, it looks like you were using a NetApp cluster configuration (which is already way better than most online communities have) and something went bad. I don't know if you're still willing to trust them after their ordeal, but setting up some sort of disaster recovery system would seem like a priority, either SnapMirror or rsync to a different system (I would recommend a server running Solaris 10 with ZFS, or the NexentaStor distribution of OpenSolaris). Obviously, the extra hardware has a cost, but I am sure the Head-Fi community will be responsive if you run a "belt-and-suspender security fund drive".