On 04/07/15 09:10, Mikael Abrahamsson wrote:

quoted

Make sure you've got your raid timeout increased - there's plenty of
threads about how to do it - otherwise one disk hiccup for any reason
is likely to cause a cascade of failures !!!!

I recommend this as minimum (in rc.local for instance):

for x in /sys/block/sd[a-z] ; do
        echo 180  > $x/device/timeout
done

echo 4096 > /sys/block/md0/md/stripe_cache_size

If you didn't do this, this could EASILY explain your problems. 7 disks
is 21TB of data. That pretty much *guarantees* TWO soft errors. Each
error will kick a disk from the array. Plus the drive you're replacing
that makes your raid 6 short by 3 drives. OOOOPPPSS.

Cheers,
Wol