As the below will show, I had a drive drop from my VM/Docker pool get out of sync. I didnt think it would be to big a job to re-add it so, I did use Ai to help get it re-added. The post below is also Ai to show what I/we did to get where we are.
I dont currently have a backup of the pool, but I am working on that.

System: Unraid (current release), 2x Crucial BX500 1TB (sdm/sdn) as BTRFS RAID1 pool named vm, holding VM disk images + Docker appdata (~700GB).

Background

One device (sdn) dropped out of the vm pool, showing as <missing disk> in btrfs filesystem show. Pool continued running degraded on the surviving device (sdm).

What I did to fix the degraded pool (all via terminal)

1. Confirmed pool was degraded with mixed raid1/single block group profiles (btrfs fi usage /mnt/vm)

2. Converted all chunks to single profile to free the missing device from RAID1 constraints:

   btrfs balance start -mconvert=single -dconvert=single -sconvert=single -f /mnt/vm
   

   (took several hours on a DRAM-less SATA SSD, ~700GB to relocate)

3. Once confirmed complete (btrfs device usage showed 0 chunks on missing device), removed it:

   btrfs device remove missing /mnt/vm
   

4. Wiped and added the replacement/same physical drive back in:

   wipefs -a /dev/sdnbtrfs device add -f /dev/sdn /mnt/vm
   

5. Converted back to RAID1:

   btrfs balance start -mconvert=raid1 -dconvert=raid1 /mnt/vm
   

6. Confirmed success — btrfs device usage /mnt/vm showed matching used values on both devices (~715GiB each), no missing devices, no warnings.

At this point, BTRFS itself considered the pool fully healthy and mirrored.

Where it went wrong

• sdn was added via CLI (btrfs device add), not through the Unraid pool GUI, so Unraid's own pool slot config never recognized it as a member.

• After an array restart, the pool wouldn't mount — Unraid only knew about sdm in its pool config, and BTRFS refused to mount a 2-device RAID1 filesystem with only 1 device supplied.

• Manually mounting both devices via terminal (mount -t btrfs -o device=/dev/sdm1,device=/dev/sdn /dev/sdm1 /mnt/vm) confirmed the filesystem and all data were 100% intact and correctly synced.

• Attempting to assign sdn to the pool via the GUI triggers "All existing data on this device will be OVERWRITTEN when array is Started" — even though sdn is a fully valid, synced RAID1 member.

• Attempting to unassign both pool devices and start with an empty pool (per the documented "make Unraid forget" procedure) fails with "Wrong Pool State — vm — too many wrong or missing devices." Pool slot count won't go below 2.

Current state (stable, not urgent, but degraded)

• Running with sdm alone assigned to the vm pool (single slot filled, other left unassigned) — mounts and starts cleanly every time, all data accessible, VMs/Docker running normally.

• sdn is fully wiped and sitting idle/unassigned — not currently part of the pool.

What I'm looking for

Has anyone hit this specific combo — BTRFS device added to a 2-slot pool via CLI, GUI refuses to recognize it (overwrite warning), AND "unassign all + restart" also fails with "too many wrong or missing devices" on a pool that requires 2 slots minimum? Trying to avoid another CLI-only balance (know it works, but takes hours and I'd rather have the GUI track this pool properly going forward). Diagnostics zip available on request.

tower-diagnostics-20260727-0933.zip

If the pool is now fixed but the GUI is out of sync, you can just reimport it (make sure you unmount it first if it was mounted manually)

on main click on the first device for that pool and then "remove pool"

back on main, create a new pool with the same name and 2 slots

assign the pool devices, leave the filesystem set to auto

start the array to import the pool

Ok will try this tonight when home thanks.