How 'critical' is /dev/md0 and /dev/md1 (OS and swap)
Hi, I have a case open with Netgear which they would like to close off, before agreeing (or not) I'd like to get a few other opinions first - there's no point in digging my heels in if nobody else would care anyway 🙂
/dev/md0 and /dev/md1 which contain the OS and swap do not appear to be monitored, that is if one of their *partitions* fails and the md becomes degraded you do not receive any alerts or anything, you just get the green light that everything is ok in FrontView
I came across this when swapping out both disks in a Pro2: - disk 1 swapped, all syncd ok - disk 2 swapped, apparently all syncd ok, FrontView reported the Volume and Disk(s) were back to 'green', everything ok ...however it wouldn't expand This was because it was in an inconsistent state, partitions not happy (from going through logs etc). I found /dev/md1 had disk 1 active, disk 2 'spare' with the state of 'clean, degraded', due to read errors from disk 1
Should disk 1 have failed at some point, the NAS would then have fallen over!
Netgear's viewpoint on this is that md0 and md1 aren't important as you can just do an OS reinstall and your data (the important bit) is still there. From my business-user point of view I totally disagree, I don't want to find out a few months after performing an upgrade that my supposed-to-be-redundant NAS is unavailable due to a single disk failure, even if 'all' that is required is an OS reinstall, reconfigure, addon setup and so on. If something is degraded, surely that's important?
