New user, came to unraid after being a sab & arr suite user on windows for a long time. Old NAS died so looked to repurpose an old PC and HDDs I had around the place. Set it all and got it going. Was working very well.

 

Added a new (old) HDD disk 5 and new SSD, adding the new drive to the array and the new SSD to the cache pool. New HDD went fine. Then noticed that I have got two thumbs down on the new drive disk 5 and additionally one of the old ones, disk 4. Didn't notice at first as the install went okay and I had carried on using the system. Maybe this caused some issues along the way while I continued using my docker apps, and it has filled up my SSD with files that are now heavily read only protected and I cannot move them. The cache pool has 1GB free and is preventing me from opening my docker apps too. Mover does nothing at all, (share is set to yes, and others set to prefer as recommended). I have also tried Dyanmix FM and that cannot delete the files either, (or indeed change permissions from owner = nobody nor can it alter the read/write access) I have tried unbalance too which although goes through the motions of moving the files from the cache, they are not removed from the SSD which remains at 118 of 120gb. SSD file system is btrfs, not sure if it should be xfs too?

 

I am in the process of shrinking the array, to remove the two HDDs flashing up thumbs down errors on the dashboard as I am wondering if those errors are causing the read only status. Unbalance worked fine here to clear some files off of disk 4 which got the thumbs down after disk 5 was added to the array.

 

Dashboard https://ibb.co/1R07SBb

cache pool devices https://ibb.co/2NSJdQb

shares https://ibb.co/myWcbfV

 

 

 

tower-syslog-20230608-2220.zip

attach diagnostics to your NEXT post in this thread

I've just run it twice now, but the browser tab keeps crashing, and the diagnostics is not completed / made available for download, can i access it on a share and retrieve it from there? Edit - found in the USB logs

 

tower-diagnostics-20230609-1610.zip

Edited June 9, 20233 yr by sihorrocks
found my own answer

Cache is missing a device and going read-only during the balance due to being full, reboot and see if you can quickly delete/move some data, if that doesn't work I can post instructions to try and cancel the balance so you can remove some data.

Thanks. I'll reboot and jump quickly into the cache drive. Will post back when I've had chance to try! Sorry a total newb so learning on the job.

Rebooted and immediately jumped into the drive / cache to delete offending files using dynamix FM, couldn't delete. Had another go with mover after disabling VMs and Docker, still can't delete the files. Definitely feel like I've tried everything

Post new diags.

Thanks!

tower-diagnostics-20230609-2102.zip

Try this:

 

-stop array

-type in the CLI/terminal:

mkdir /x
mount -o skip_balance /dev/sdi1 /x
btrfs balance cancel /x
umount /x

-re-start array and post new diags.

Got this in the CLI

root@Tower:~# ^[[200~mkdir /x
bash: $'\E[200~mkdir': command not found
root@Tower:~# mount -o skip_balance /dev/sdi1 /x
mount: /x: wrong fs type, bad option, bad superblock on /dev/sdi1, missing codepage or helper program, or other error.
       dmesg(1) may have more information after failed mount system call.
root@Tower:~# btrfs balance cancel /x
ERROR: not a btrfs filesystem: /x
root@Tower:~# umount /x~
umount: /x~: no mount point specified.
root@Tower:~#

 

screwed up there, and tried again to enter in the cli

root@Tower:~# mkdir /x
mkdir: cannot create directory ‘/x’: File exists
root@Tower:~# mount -o skip_balance /dev/sdi1 /x
mount: /x: wrong fs type, bad option, bad superblock on /dev/sdi1, missing codepage or helper program, or other error.
       dmesg(1) may have more information after failed mount system call.
root@Tower:~# btrfs balance cancel /x
ERROR: not a btrfs filesystem: /x
root@Tower:~# unmount /x
bash: unmount: command not found

 

tower-diagnostics-20230609-2138.zip

Edited June 9, 20233 yr by sihorrocks
additional info

Try this instead:

mkdir /x
mount -t btrfs -o skip_balance /dev/sdi1 /x
btrfs balance cancel /x
umount /x

If the mount command still fails don't continue, just post the output

yeah still failing

root@Tower:~# mkdir /x
mkdir: cannot create directory ‘/x’: File exists
root@Tower:~# mount -t btrfs -o skip_balance /dev/sdi1 /x
mount: /x: wrong fs type, bad option, bad superblock on /dev/sdi1, missing codepage or helper program, or other error.
       dmesg(1) may have more information after failed mount system call.
root@Tower:~#

 

Also error notification on gui from the cache drive -

 

cache pool btrfs missing device(s) <drive name>

Sorry, forgot that the pool is missing a device, so it can only be mounted degraded:

 

mkdir /x
mount -t btrfs -o skip_balance,degraded /dev/sdi1 /x
btrfs balance cancel /x
umount /x

 

worked this time, great, no errors in CLI - array restarted, new diag attached

 

heres the CLI inc typo

 

root@Tower:~# mount -t btrfs -o skip_balance,degraded /dev/sdi1 /x
root@Tower:~# unmount /x
bash: unmount: command not found
root@Tower:~# umount /x
root@Tower:~#

 

tower-diagnostics-20230609-2231.zip

Edited June 9, 20233 yr by sihorrocks

It's still crashing while trying to delete the missing device, I think the only option is to mount the pool read-only, backup data and re-format, you can mount read-only with:

mkdir /x mount -t btrfs -o skip_balance,degraded,ro /dev/sdi1 /x

 

Pool data will be under /x, you can then back that up somewhere else.

 

 

Thanks for your time on it, much appreciated. I might back the data up via the windows share to another device and nuke it, start again, copy it over. Bummer.

Those SSDs are probably going to be too small for both storing docker/VM files and also for caching user share writes. Unless you write very little each day maybe even too small for just caching. You should consider larger SSDs.