2018-05-24 04:39 AM
Dear NetApp community,
It so happened that I have to recover a NetApp E2724 storage after some, well, to say it mildly, "not quite careful" operations were performed on it by the previous administrator.
I'll try to briefly explain what the situation is.
There is a 24SFF-bay case. It used to host 16 10K Hitachi drives with RAID6 disk pool created. As I was told, at some point in time 4 more drives were added to the empty slots. Those drives turned out to be different (HGST reported as 10,52K). As I was told, the storage at that point informed that it needs at least 12 identical disks to create a disk pool and therefore the operation was aborted. Those 4 drives just stayed there uninitialized and not added to a disk pool (or so was believed) for quite a while.
Fast froward several months. Eight new Hitachi drives (just like those initial 16 ones) arrived. The admin assumed that those four incompatible are not part of anything and are unused so proceeded to remove them. When he did, everything on that 16-drive pool went down with "Status: failed". He then proceeded to insert the disks back into their slots with a hope that it will acknowledge them and reassemble the RAID. unfortunately it did not.
Now, to the main question. Tl/dr: There was no failed drive. Four drives from RAID6 disk pool (that should not even be in that disk pool, but that's not important) were removed and then inserted back again. Status of that disk pool is "Failed". But the data are physically still on the drives. How can I reassemble the array after such incident?
Much thank you in advance for any assistance!
2018-05-24 07:41 AM
@MrkF I would strongly recommend that you file a support request with NetApp support. When it comes to data recovery, far better to be safe than sorry. While there probably are some individuals on here that could tell you the right commands to run, NetApp support will know the safest way to recover the pool.
2018-05-28 02:36 AM
Thank you for your answer. At this point data safety is of low importance because everything has been already successfully recovered from backups. That said I'd still like to try rebuild that disk pool as an exercise and a chance of getting some valuable experience. This situation just doesn't make any sense to me, even a consumer level RAID controller in my desktop can easily recover from member drive hot unplug.
As I understand, the revive command is only used with volume groups. How should I approach a disk pool then?