Paul Brook wrote:
I have the following configuration: a thecus n2100 with two disks configured as raid1 (actually there are three partitions on each disk: a non-mirrored swap partitionRookie mistake. If either drive goes down you suddenly loose swap, and the whole machine is foobared.
Uhm, you are obviously right: let's say that when I partitioned the disks I didn't have swap space resiliency at the front of my thoughts.
The swap partition on the dead disk actually had some swap in use (48 blocks, per /proc/swaps) but swapoff seems to be capable to stop using it or at least it doesn't complain and the dead partition disappears from /proc/swaps.
But still no reboot. Do you think it is really messed up (I don't understand how the process(es) those 48 pages belonged to can survive without them) and there is no hope of a clean shutdown?
linux will do load ballancing over multiple swap partitions, however it does not provide any redundancy.Paul