VM/Docker Reports Disk is Full - Cache Drives Disappear on Reboot - Full Reboot Sorts The Issue?

[mrforsythexter]

So, I have been having a strange issue.

 

Every other day or two, I start getting messages from VM's and Docker Containers, They are full, or they stop. No updates can be installed for the dockers containers.

 

Because I just didn't have a clue, I restarted the machine from the web interface, it comes back up, but the array is offline, and the cache disks are missing (please remember they are showing as fine before the reboot).

 

The cache drives look fine, reporting 381GB Free.

So, I shutdown the server from the web interface and turn it back on via the button. Everything comes up and is fine for days, then it happens again.

 

It's like the controller for the cache drives fails or something, but unraid is unable to tell this has happened.

 

ls /mnt/cache/
/bin/ls: reading directory '/mnt/cache/': Input/output error

 

The drives are still listed via `lsblk`

I can't stop the array (via web interface), the other disks (array)(Spinning rust) are accessible and writeable on the same sata controller.

 

Any advice or direction would be appreciated.

tower-diagnostics-20221019-1921.zip

[JorgeB]

Both cache devices dropped offline, one at a time, replace cables and run a scrub, see here for more info.