I have a weird case last week.

 

My unraid pool is set up with 1 parity + 3 disks. There is extra one cache disk

Share "CCTV" is set up in the pool, with cache set to "YES".

I am running docker ShinobiCCTV with recording set to /mnt/user/CCTV

So the recording will be saved in cache disk, with mover run every night.

 

Disk 2 on pool is having bad sector, and I decided to RMA the disk. I removed Disk 2, and run the pool from parity drive with missing Disk 2.

The ShinobiCCTV Docker decided that it cannot find the CCTV share path, resulting in crash. When I changed the recording location map to /mnt/user0/CCTV thus bypassing the cache it worked fine.

 

The new replacement disk came, restored the content, and set back to /mnt/user/CCTV and everything runs normal again.

 

Did I do something wrong, or with missing disk, dockers cannot access the cache drive?

It should work normally with the emulated disk, there could have been a file system corruption issue, but diganostics from the time would be needed.

2 hours ago, philliphs said:

unraid pool is set up with 1 parity + 3 disks. There is extra one cache disk

we usually refer to this as "the array" and reserve the word "pool" for Unraid storage assigned outside the array, cache, for example.

On 2/25/2023 at 6:49 PM, JorgeB said:

It should work normally with the emulated disk, there could have been a file system corruption issue, but diganostics from the time would be needed.

Yes. Emulation was supposed to happen, but it didn't for that specific docker. I suppose it had something to do with the cache system.

 

I will dig a bit more when I'm replacing another of my ironwolf soon. Funny I have mixed disks of WD red and Ironwolf, I had bad sectors on 3 Ironwolfs (RMA was easy) but none on WD red.

Edited March 11, 20233 yr by philliphs