I have an array of 3 12TB disks, one of which is parity.

 

I was copying 3 zip files of Google Takeout pics totaling 110GB from a USB drive. Seems like it stopped writing to the data disk that I was copying to but continued to copy to the emulated drive. I saw the error after the copy was complete. When I unassigned disk2 and mounted it in Unassigned Devices, I could see that all my previous data is there, including the 2 50GB files that I was copying. Copying was interrupted during the last 10GB file and only around 7GB were copied.

 

I removed the disk physically and mounted it on my PC and see no read or write errors in SMART stats. I've now changed the data cable and the SATA port that the disk2 is connected to on the motherboard. Last parity check was done 3 days ago without issues.

 

How should I proceed, knowing that only 1 file is bad on the disk. According to documentation, I should use the "Rebuilding a drive onto itself" option. But would it not try to rebuild all of the 12TB? I guess it would be good because if there is a real issue, it will show up during the rebuild. Is there another way to proceed?

 

I'm attaching some syslogs and disk logs along with screenshots.

 

Screenshot of Array

 

SMART stats of disk2

 

Actual contents of disk2 (after copy was complete)

 

Emulated contents of disk2  (after copy was complete)

 

 

syslog to share.txt Disk Log to share.txt

Edited December 27, 20241 yr by New to Unraid User
adding info between images

From the UnRaid web admin run a short SMART test and pull the logs from the drive. 

That model of hard drive is not for consumer sale/use.  I'd contact Seagate or search for UNRAID HSMR before moving forward.

Exos X14 - SATA 512E/4kn SED 12TB ST12000NM0558 P/N: 2LL111-268

I've been using these drives for a little more than a month on Unraid. I think you may be getting confused with HA-SMR not being compatible with Unraid?

These are Hybrid SMR drives, not HA-SMR. The Hybrid SMR drives are shipped from Seagate with CMR as the default and have the ability to be converted to SMR. These were not made for public. The 12TB capacity is the CMR capacity. 

Assuming the emulated disk is still mounting and contents look correct, replace the cables and try rebuilding on top, then look in the syslog for more ata errors during the rebuild:

 

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

Rebuild was successful with no ATA or other errors reported during the rebuild. I had stopped all dockers and VMs during the rebuild. I also stopped the server and mounted Disk2 to verify contents.

I'm hoping it was just a bad connection. Anything else I should keep a lookout for?

Notice [SERVERUNRAID] - Data-Rebuild finished (0 errors)
Duration: 17 hours, 48 minutes, 26 seconds. Average speed: 187.2 MB/s

 

Should be resolved, just keep monitoring the log for anymore ata errors during future checks or high i/o.