On 25/05/2020 23:18, Thomas Grawert wrote:
Okay, so we need to grep dmesg looking for a message like that above
about reshaping.
So let's grep for "md/raid" and see what we get ...
root@nas:~# dmesg | grep md/raid
[ 321.819562] md/raid:md0: not clean -- starting background reconstruction
[ 321.819564] md/raid:md0: reshape_position too early for auto-recovery
- aborting.
(again: thanks a lot for your help. I never expected that! :-) )
Okay, so go back to the previous email for an example of how to revert
the reshape - it's done exactly what I expected and the reshape has
failed to start.
Once you've reverted the reshape, it should start fine. And with an
up-to-date mdadm and kernel, it should start and finish reshaping fine.
So step 1, revert the reshape. Step 2, get the array back running. Step
3, start the reshape again.
Cheers,
Wol
