As a belt and suspenders approach, I also do daily backups of key data to a large ATA store on my backup server. The backup server can then do a local backup to tape without additional network traffic.
Double drive failures are rare, but one failed drive can bring down the SCSI bus. I have had that happen to me and also had failures of the hotplug backplane.
Depending on the robustness of the server, there are a raft of failures that could necessitate more than a drive replacement. Anything from power supplies, to RAID controller could complicate things. On my mission critical systems I even have hot spare servers.
Belt AND suspenders AND ductape?
_________________________
Give a man a fish and he will be back for more. Slap him with a fish and he will go away forever.
