Hey all,

Recently got a new 1tb SSD for my cache. Thought I could add it to my existing cache pool with a 2tb drive.

Since then, my intial cache drive has started displaying as full at 1.31tb. The drive appears to be loading in read-only mode - mover doesn't move files, and scrub aborts after 0 seconds. btrfs filesystem usage displays the following:

Overall:
    Device size:		   2.75TiB
    Device allocated:		   2.56TiB
    Device unallocated:		 199.12GiB
    Device missing:		     0.00B
    Device slack:		     0.00B
    Used:			   2.37TiB
    Free (estimated):		 244.14GiB	(min: 217.15GiB)
    Free (statfs, df):		 128.00KiB
    Data ratio:			      1.57
    Metadata ratio:		      2.00
    Global reserve:		 512.00MiB	(used: 16.00KiB)
    Multiple profiles:		       yes	(data, metadata, system)

                  Data      Data      Metadata Metadata  System    System                              
Id Path           single    RAID1     RAID1    DUP       RAID1     DUP      Unallocated Total     Slack
-- -------------- --------- --------- -------- --------- --------- -------- ----------- --------- -----
 1 /dev/nvme0n1p1 708.01GiB 951.84GiB  2.00GiB   2.00GiB  32.00MiB 16.00MiB   199.12GiB   1.82TiB     -
 2 /dev/nvme1n1p1         - 951.84GiB  2.00GiB         -  32.00MiB        -     1.02MiB 953.87GiB     -
-- -------------- --------- --------- -------- --------- --------- -------- ----------- --------- -----
   Total          708.01GiB 951.84GiB  2.00GiB   1.00GiB  32.00MiB  8.00MiB   199.12GiB   2.75TiB 0.00B
   Used           522.99GiB 951.84GiB  1.61GiB 271.78MiB 208.00KiB 32.00KiB       

I've been attempting to copy data off the cache to the array (in case I have to wipe the thing), and am noticing btrfs errors:

Sep 24 16:48:35 NORTHDAMEN kernel: BTRFS error (device nvme0n1p1 state EA): level verify failed on logical 59627181359104 mirror 1 wanted 0 found 1

Sep 24 16:48:35 NORTHDAMEN kernel: BTRFS error (device nvme0n1p1 state EA): level verify failed on logical 59627181359104 mirror 2 wanted 0 found 1

I've ran a file system check and receieved no errors. Pool device stats also show no errors.

I'm a little unsure what to do next - I have appdata backups so I can always wipe and restore, but I'm hoping that is a last resort (backup is about three weeks out). Wondering if there's a way to revert the cache pool to just use the 2tb drive? Will take any advice.

Thanks,

Edited September 25, 2025Sep 25 by frigginbrownie84

Default is for raid1 btrfs mirror. The pool capacity is equal to the smaller of the 2 drives.

https://docs.unraid.net/unraid-os/using-unraid-to/manage-storage/cache-pools/#changing-pool-raid-levels

You probably want to change it to Single Mode.

Thanks - I've tried converting it to single mode via the GUI but nothing seems to happen. I have noticed the following warnings appearing:

pool BTRFS too many profiles (You can ignore this warning when a pool balance operation is in progress)

Is there another method to change back to single mode?

Post new diags after trying to convert it.