During a parity check, my office (and HBA card) got very hot (A/C went out).

Now I have two drives that are disabled in my array. I'm running two Parity drives so my data I hope is safe but for example, Plex no longer loads even though the log says it loaded successfully. I don't think the drives are bad, but how do I re-enable them?

 

Running 6.12.10 and attached the syslog and SMART logs for the two disabled drives.

syslog.txt WDC_WD120EDAZ-11F3RA0_8HGY00SY-20240804-1515.txt WDC_WD120EDBZ-11B1HA0_5QG88LTF-20240804-1513.txt

Please post the diagnostics with the array started.
 

My apologies. Restarted the array and the Diagnostics. 

elwood-diagnostics-20240804-1823.zip

SMART looks OK, so most likely a power/connection issue, since the emulated disks are mounting, and assuming contents look correct, you can rebuild on top, both at the same time, I would recommend checking/replacing the cables first to rule that out:

 

https://docs.unraid.net/unraid-os/manual/storage-management#rebuilding-a-drive-onto-itself

 

 

Following up with this to give you some results.

 

I attempted to rebuild both drives at once using the instructions above. One drive failed out with errors and was disabled and the rebuild of the other seemed to stop. Let it run overnight with no further progress. 

 

Rebooted and removed and re-added the drive that did not show errors on the first rebuild. Took a day but went through with no errors. So I thought I'd try the other that had errors. It's currently at 25% with no errors where it errored out at just 8% the first time. 

 

I'm starting to think the LSI card is going bad because: 1. this all started with a parity check throwing errors, 2. rebuilding two drives at the same time threw errors but one at a time doesn't show any errors. 

 

Thoughts?

If there are new errors post the diags before rebooting. 

Data rebuild complete on both drives, but something went awry as now my Plex has config as lost the configuration for movie and tv show shares.

 

Doesn't look like SMART errors on rebuild.

 

 

elwood-diagnostics-20240805-1829.zip

There are constant power on errors with multiple disks, and during the rebuild there were read errors on a different disk, so the rebuild will be corrupt.

 

These errors look more like a power/connection issue.

It might be time for an upgrade. I currently have a 700W thermaltake running an i5-9400F, Quadro P2000, 2 LSI HBA cards, with 14 drives. Throwing all of the drives into usage at once might be driving voltage drops. 

 

Is there a way to fix the corruptions that have already taken place? I'm seeing a suggestion in the log to "Unmount and run xfs_repair"?

You should run xfs_repair (check filesystem without -n), but some data may remain corrupt, depending on what the read errors affected.