Forum Discussion

Aspirant

Feb 26, 2022

628X reseated disk not failing back

For some reason, one of the disks in my readynas was reported failed. So it started rebuilding on the hotspare. I reseated the supposedly failed disk, and instead of promptly failing back from the ...

Sandshark

Sensei

Feb 26, 2022

jcdick1 wrote:

How do I get the ReadyNAS to appropriately restore the now-fine reseated disk to the raid set and return the hotspare to hotspare status?

You can't till the current re-sync with the spare completes, and you shouldn't have tried. Once the NAS dropped the drive out of the RAID, it can't be directly added back in because it was out of sync. The NAS did exactly what it;s supposed to do when there is a global spare, replace the failed drive with it and start a sync. and you should have just let it finish doing so. But if you didn't have a global spare, it still wouldn't have just put the drive back in the RAID set, it would have to do a re-sync with that drive.

Just removing and replacing a drive is not usually a good plan in any case, anyway. The NAS kicked it out of the array for a reason, and it's best to figure out why before you either pull and re-seat it to let it re-sync or replace it if it's bad. Using vendor tools for the drive in a PC, either connecting the drive directly or via a USB dock is best for doing that. But looking at the NAS log for SMART errors is also useful. You say the drive is "now fine" and I suspect you don't really know that since the NAS isn't even trying to use it

Your NAS is reporting a second volume because it sees, but doesn't know what to do with, an old fragment of the volume on the re-inserted drive. I recommend you let the sync with the spare complete before you do anything else and potentially make matters worse.

Once the sync is complete, remove the drive that orignally failed and test it with vendor tools. If it really is OK, and you really want to replace the previous spare with it, that's exactly what you need to do. It would be best if you remove the partitions from the drive first with a PC, then remove the previous spare from the NAS and replace it with the old drive. Once you've added it back to the array, let it re-sync. Then remove the partitions from the old spare, put it in, and make it a global spare. Or if you don't want this to happen again, keep it on the shelf as a cold spare.

jcdick1
Aspirant
Feb 26, 2022
Okay, I guess. With our other storage (Dell/EMC and HP DL-class storage servers), the first thing done if a drive fails is just reseat it. The sync to the hotspare immediately stops and resyncing to the reseated drive begins and *then* its determined if the drive is actually bad. Because the system is intelligent enough to read the disk metadata and say "Oh, hey, I know you!" And a resync is faster to the reseated drive because the data is mostly already there.

But I'll just pop out the drive again and wait four or five days for the rebuild to complete. Then wait another four or five days for it to do it again. Doesn't seem like it should be that way, but I guess its how it is.
- StephenB
  Guru - Experienced User
  Feb 27, 2022
  jcdick1 wrote:
  
  And a resync is faster to the reseated drive because the data is mostly already there.
  
  FWIW, not the case with the ReadyNAS. The system does not assume anything about the data on the reseated drive - the resync reconstructs every sector from the remaining drives in the volume.
  
  jcdick1 wrote:
  
  But I'll just pop out the drive again and wait four or five days for the rebuild to complete. Then wait another four or five days for it to do it again.
  
  If you want to ensure that the inactive volume issue won't pop up again, then unformat the drive while it is out of the NAS. If you are testing it while it is removed (which I do recommend), then the test that writes zeros to the drive will of course unformat it for you.