Aside from the interruption to their business, the carnage that was & still is the Christchurch CBD and very stressed staff, there is a little IT silver lining for them.
We've been able to recover all their older servers (some physical, some virtual) to a new freaky-fast HP DL380 with lots and lots of SAS 15K drives in it. And this has been a golden opportunity to upgrade lots of their bits of software (latest AV client, latest anti-spam product, latest version of Citrix, etc etc) that have been on the cards for a while but they didn't have the time for downtime.
At the end of this they're going to have way faster servers, gigabit to the desktop & tons of free space on their servers.
As the horse has well and truly bolted we're going to close the gate this time with disk imaging backup software and take that to tape from a SAN instead of backing the servers themselves up to tape directly. Much faster to recover in the event of another disaster, and a lot less painful for the engineer trying to do it.
One cool thing I've found - VMWare ESX 4.1 update 1 can physically assign a PCIe SCSI card directly to a Windows Server 2000 VM - didn't know that could be done until I needed to access their new tape drive and found that I couldn't pass the tape drive directly through.
VMWare's probably been able to do this for a while, I just didn't know about it until now. I also didn't know that ESX 4 couldn't address a LUN bigger than 2TB - spent an hour or so wondering why my 2.045 TB array was showing up as 0 bytes in size by ESX when all the components were on the HCL. Just needed to resize the volume to just under 2TB and away it went. Lesson learnt.
Other related posts:
Extending the CEPH cluster, things we've learnt
Creating redundant, clustered & scalable storage - a DIY guide
Building A Win8.1 based Chromebook - A How To