I've been asked to implement a rather strange DR setup and I could use some advice - I don't think it's the way to go but that's another story.
I have 3 main physical servers which are currently backed up to a NAS disk - full backup of data takes place on a Sunday as does a full backup of Exchange 2003. Incrementals for data and Exchange run on the days in between full backups.
The plan is to use VMWare to create 3 VMs of those servers on a dedicated DR server. The DR server will live off site. Periodically a 1TB passport drive will be used to capture fresh images of those 3 servers which will then be taken to the DR server off site and the 3 VMs will be replaced by the new images.
That seems reasonably OK. However, to restore data to one or more of the VMs should a physical server die between VM image captures the plan is to use the BExec 2010 data I backup from the physical servers e.g. the VM of our Exchange server would (in theory) have more recent full and incremental backups restored to it from my NAS B2D files. The updated VM server would then stay virtual.
I'm trying to test this by copying the most recent full Exchange backup to a USB disk and importing that in to the VM of our existing BExec server (which is also one of the 3 main servers that need a DR solution). The person who dreamed this up then expects to carry out a restore from the latest BExec data (now on the VM of the backup server) to the VM of our Exchange server i.e. carrying out the Exchange restore from within the VM environment.
In preliminary tests BExec on the virtual backup server does actually start to chew the restore job but after a while I see strange error messages appear on the server, it throws an exception and a forced shutdown takes place. Can someone suggest if this is a sensible way forward, if there is a fundamental flaw in the plan or (if there isn't anything wrong with the plan) why BExec dies part way through the Exchange restore?
Thanks in advance.
