Btrfs error Unmountable: No file system on cache

[Magarac]

Hi there

 

First of all, I feel somehow stupid in what I did:

 

When I setup unraid I moved the appdata, system and domains to the cache disk(cache only) so that my array can go to sleep/spin down while it's not used.

 

My cache is a nvme 1tb and an ssd 1tb...it automatically was joined to a raid 1.

 

Now after some months my SSD was throwing saome failures and I couldn't write on cache (fatal error in windows and plex database only in read-only.

 

Now where it beginns: Instead of making a backup of all these shares and file, I wanted to rebalance the cache for single use.

 

I clicked on my working nvme and said single use and balance....during the balancing I had to restart it several times because the SSD wasn't responsive anymore.

 

After the balancing was finshed the sdd was still member of the Pool?? then I stopped the array again and removeed the ssd. Since then I have the error on the cache that "Unmountable: No file system on cache"

 

I tried several things from the community and faq but still no luck.

 

Can anybody please guide me? I wished I could restore some data like the system share and appdata share.....otherwise I don't know how to newly create these shares.

 

Many thanks for your help already

[Magarac]

please tell me if I can upload something more, thanks

[JorgeB]

 

This suggests the device was wiped/trimmed, only option would be a file recovery program, like UFS explorer, if the data wasn't discarded by the SSD.

[Magarac]

Thanks JorgeB, I will try do something with usf explorer....but if this doesn't work? How can I bring these shares running again because now I don't even see in "Shares" these folders anymore.

 

I feel really stupid about this error....

[JorgeB]

Shares should still work, but we'd need the diagnostics to say more.

[Magarac]

attached you can see the daignostics and also the syslog.

tower5-diagnostics-20210912-1248.zip tower5-syslog-20210912-1048.zip

[JorgeB]

With array started please.

[Magarac]

didn't know it makes a difference. sorry. Attached the logs with started array

tower5-diagnostics-20210912-1300.zip tower5-syslog-20210912-1100.zip

[JorgeB]

User shares should be working, /mnt/user is mounted, but since the pool isn't mounting unassign al cache devices and start array to see if it makes a difference.

 

P.S. syslog is included with the diags, no need to post it separately.

[Magarac]

I stopped array, put cache on "no devices" and started array:

New Diagnostic attached.

 

libvirt/vmmanager and so on still no start

 

Thanks for the hint, diagnostics is then the all-in-one log:)

tower5-diagnostics-20210912-1313.zip

[JorgeB]

10 minutes ago, Magarac said:

libvirt/vmmanager and so on still no start

I was looking at shares, not services, system was likely on cache, you need to re-create on the array, but note that docker and VMs will be empty, dockers can be easily restored if you have an appdata backup, for VMs you need backups of the vdisks and libvirt.img.

[Magarac]

so in facts I can just create new shares for appdata/domains and system, redownload libvirt and the docker image and then I'm fine again? of course just lost data but system is ready and good to work again?

[JorgeB]

Correct.