Urgent - Cache drive corrupt (pool)

[Seriously_Clueless]

Hi all,

 

Here is what let to a corrupt cache drive on my Unraid installation:

 

I have two cache drives, 1x1tb and 1x 2tb. Since they are mismatched I ended up buying another 1tb to replace the 2tb.

 

So I googled and followed this post to the . https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=511923

 

It worked out fine but the pool size was 1.5TB even after the drives swapped. I moved all files to the btrfs cache again without any error message. I was a bit wondering since the cache 1 drive was the only one that had writes to it, not the cache 2 drive even so they are in a RAID1. Then I stopped the array after it was finished and restarted unraid.  After the restart the cache drives is showing:

 

root@Tower:~# btrfs check --readonly /dev/nvme1n1p1
Opening filesystem to check...
bad tree block 559374450688, bytenr mismatch, want=559374450688, have=0
Couldn't read tree root
ERROR: cannot open file system

root@Tower:~# btrfs rescue super-recover /dev/nvme1n1p1
All supers are valid, no need to recover
root@Tower:~# btrfs rescue chunk-recover /dev/nvme1n1p1
Scanning: DONE in dev0                         
open with broken chunk error
Chunk tree recovery failed

 

Thank you in advance for any advise. Please let me know if there is any way I could recover the missing files.

 

Hopefully I can recover the cache since I really would hate loosing all my docker and vm data.

 

Thank you.

[JorgeB]

See here for some recovery options.

[Seriously_Clueless]

Thank you, will give it a try.

[Seriously_Clueless]

Unfortunately none of the suggestions yielded any benefits. I had to wipe the cache drive. What a bummer especially since I tried following guidance doing it in the first place. 

 

The new Raid 1 cache pool is now created. How can I backup my cache drive pool on a regular basis so if this would happen again I have a perfect copy of it on my array? 

[JorgeB]

There are various possibilities, you can use snapshots if it's btrfs or just something like rsync.

[Seriously_Clueless]

Thanks but would a rsync of a running array cache drive provide a usable backup? I am concerned that i.e. the dockers might run and then the backup might not be usable afterwards?

 

Do I have to backup each vm and each docker separately and ensure they are offline when doing so?

 

Thank you. If there is a link that explains this in more detail, happy to read if anyone can point me in the right direction.

[JorgeB]

To use rsync you'd need to stop the VMs first, docker should still be OK, since you don't need the docker image, only appdata, or use the CA appdate backup plugin witch automatically stops then restarts them, alternatively and for everything you can use snapshots.