Suppose you have a server with a hardware raid controller configured for Raid 1 (so if one of the hard disks goes bad, the other essentially acts as failover until you can replace the failed disk).
I know the reasoning for using multiple partitions in linux is so that if the filesystem goes bad then it only breaks part of your system (let's say /var) rather then the entire system.
I always imagined that this was meant specifically for hardware failures, which the raid 1 should prevent.
Is it safe to put everything under one partition (plus swap space of course) when using a raid 1 array, or are there other software issues relating to the filesystem (ext3) which can occur even if hardware failures can be prevented?
Thanks for you time