[BlueOnyx:17583] Re: RAID1 failure BlueOnyx 5208R

[Tigerwolf]

On Tue, 12 May 2015, Tom wrote:

> cat /proc/mdstat
> Personalities : [raid1]
> md0 : active raid1 sdb1[1]
>      511988 blocks super 1.0 [2/1] [_U]
>
> md1 : active raid1 sdb2[1]
>      243625852 blocks super 1.1 [2/1] [_U]
>      bitmap: 2/2 pages [8KB], 65536KB chunk

It looks like it's running with only one drive active:  sdb  using 
partitions 1 and 2.

So it's indeed 'degraded' since one of the drive pair (presumably sda) is 
unused and is totally out of the picture.

Check the drives with smartctl -a /dev/sd(x) to see if there's an obvious 
drive failure issue that may be the cause.  Also, look through logs to see 
if there was some problem noted.

If the unused drive looks ok, make a good *external* backup of the entire 
system, then try re-adding the missing drive partitions with
       mdadm /dev/md(partition) -a /dev/sd(partition)

for each of the two partitions.  This will attempt to rebuild the raid 
array and re-mirror the data from the one functioning drive onto the 
unused one.  For a large drives/partitions, this will take quite some time 
to copy all the sectors (and the machine will slow down accordingly).

I have seen systems that shutdown uncleanly decide upon reboot that one of 
the raid drives 'isn't worthy' and will activate only one of the pair. 
I've also seen a failing drive get kicked out of the array.

Regardless, of the cause, and even if the rebuild is successful, regard 
the machine suspiciously until you're satisfied that it's functioning 
normally.