Need Guidance - Incomplete Cache Drive Replacement

[a12vman]

I noticed a few errors on my NVME drive(holds System,Docker,AppData,Domains) after a recent 6.12.6 upgrade.. I picked up a new SSD, formatted it & added as a Cache Drive.

 

Last night I stopped docker & VM services.  Went to System,Appdata,Docker,Domains shares & changed secondary storage to Array, Mover action set to Cache-->Array.  Kicked off the Mover & let it run overnight.

 

This morning I confirmed the mover wasnt running.  Went to the same shares, changed primary storage to Crucial_Cache, set Mover Action to Array-->Cache.  Started the Mover.  Within minutes it failed. sdi is the new cache drive and filled the system log.

 

When I look at Shares, Appdata and System are Green and living on Array.  Docker and Domains show that files are unprotected.  Looking at MC I have Docker.img on Mnt\User\Docker.  Domains is still on the NVME Cache even though I set action to NVME-->Array.

 

Am I incredibly unlucky with a bad SSD or could I have a problem with the controller?  What are my next steps? I don't want to make the problem worse.  I do have a full backup of AppData that I took before I started all of this.

 

 

Dec 30 06:12:07 MediaTower kernel: BTRFS error (device sdi1): bdev /dev/sdi1 errs: wr 6, rd 0, flush 0, corrupt 0, gen 0 Dec 30 06:12:07 MediaTower kernel: BTRFS error (device sdi1): bdev /dev/sdi1 errs: wr 7, rd 0, flush 0, corrupt 0, gen 0 Dec 30 06:12:07 MediaTower kernel: BTRFS error (device sdi1): bdev /dev/sdi1 errs: wr 8, rd 0, flush 0, corrupt 0, gen 0 Dec 30 06:12:07 MediaTower kernel: BTRFS error (device sdi1): 

mediatower-diagnostics-20231230-0643.zip

Edited December 30, 2023 by a12vman

[JorgeB]

New device dropped offline:

 

Dec 30 06:12:07 MediaTower kernel: ata5: reset failed, giving up
Dec 30 06:12:07 MediaTower kernel: ata5.00: disable device
Dec 30 06:12:07 MediaTower kernel: ata5: EH complete

 

This is usually a power/connection issue, but since it's an MX500 also check it's using the latest firmware, since it's a known issue with some older ones:

 

https://forums.unraid.net/topic/137795-6115-issues-with-btrfs-cache-after-ssds-replacement-and-hw-upgrade/?do=findComment&comment=1265949

 

[a12vman]

Can I change Docker Vdisk & Storage Location to run off the array for the time being?

 

[trurl]

Yes. After you get the pool working again, simply changing those settings will not move anything though.

[a12vman]

Ok my biggest concern is preserving my plex metadata.  I changed the docker vdisk and appdata location as below.  I enabled docker.  Plex now shows as empty.  In looking at the filesystem the original plex appdata on the array is unchanged at 4.5GB.  It re-created the appdata on cache_nvme as a new install only 4 MB. The other Dockers like Emby & Sonarr have the same issue but others like Deluge & Duplicacy are ok. What did I do wrong?  

Edited December 30, 2023 by a12vman

[trurl]

My guess is you started the container when its previous appdata was not accessible at the configured path, so it created new appdata.

 

You said you had backup appdata.