You don't, pools are a single filesystem comprised of all disks (original RAID-type setup), the UNraid array is what's special with each disk being a separate filesystem.

From what I can see you were told a couple of weeks ago how to run btrfs commands to remove a disk from the pool.

1 hour ago, Moker-Vet said:

. i cannot see one disk, its just the entire storage pool so i cannot move the data of one disk to another and exclude it from using. how would i overcome this?

You cannot.

If your pool is redundant, then you could copy files off the pool and remove one disk. Failing that it is case of getting all files off the pool before removing any drives from the pool.

4 hours ago, Kilrah said:

a couple of weeks ago

The slow pace of this is probably due to very out-of-sync timezones. I look at something and comment and ask for more, then many hours later a response, then many hours later I respond, etc. Also, I expect it to take more weeks simply due to the large amount of data and size of disks involved.

4 hours ago, itimpi said:

If your pool is redundant

btrfs raid5 so we are going to work on getting pool drives moved to the array one at a time without having to put all 86TB of the pool data somewhere.

The most recent diagnostics showed /mnt/user0 (the array) had 36T free space. That should be enough to allow us to free up one of the 22TB drives from the pool so we can use it as parity in the array.

5 hours ago, Moker-Vet said:

have set all shares (except moker-vet) that exist on moker-vet pool to move to array, this includes:

backup

downloads

media-landingzone

synology-nas

Lets go ahead and move those then, that will take care of 6TB, then we can consider how to take the 80T moker-vet share in smaller bites to move off the moker-vet pool.

Post new diagnostics after you have moved those other shares to the array.

16 hours ago, trurl said:

The slow pace of this is probably due to very out-of-sync timezones. I look at something and comment and ask for more, then many hours later a response, then many hours later I respond, etc. Also, I expect it to take more weeks simply due to the large amount of data and size of disks involved.

yeah this is what i figured to, also i have a busy job so i have a few hours each day being able to respond, apologies if this is more time consuming this way.

16 hours ago, trurl said:

btrfs raid5 so we are going to work on getting pool drives moved to the array one at a time without having to put all 86TB of the pool data somewhere.

The most recent diagnostics showed /mnt/user0 (the array) had 36T free space. That should be enough to allow us to free up one of the 22TB drives from the pool so we can use it as parity in the array.

thats great, lets work towards that

16 hours ago, trurl said:

Lets go ahead and move those then, that will take care of 6TB, then we can consider how to take the 80T moker-vet share in smaller bites to move off the moker-vet pool.

Post new diagnostics after you have moved those other shares to the array.

its in progress! i will post when its done

On 2/19/2026 at 9:43 AM, trurl said:

The most recent diagnostics showed /mnt/user0 (the array) had 36T free space. That should be enough to allow us to free up one of the 22TB drives from the pool so we can use it as parity in the array.

And moker-vet pool already has 59T free, which will go to 65T after you move those other shares to the array.

1 x 22T will become parity, so that would only leave 43T free on the pool. We need to leave the other 22T there for now so it can become parity2 later. So, we could also remove 2 x 20T from the pool, giving 40T to add to the array free space. That only totals to 76T free on the array though, so not quite enough to move remaining 80T mokervet share from the pool. We will still have to move that share in 2 smaller chunks.

Hi Trurl,

A in between update, i managed to purchase 1 more 22tb hdd disk. i will try to get one more but that might take some weeks, this will relieve the burden abit i suppose. this does mean that i have to replace another disk for the 22tb one. i will try and replace a 2tb and a 4tb disk for these 2.

4 hours ago, Moker-Vet said:

purchase 1 more 22tb hdd disk

That could be parity instead of the one we were planning to remove from the pool to become parity.

The large disks we remove from the pool will replace the smaller array disks.

The immediate actions we had discussed:

On 2/19/2026 at 9:43 AM, trurl said:

Lets go ahead and move those then, that will take care of 6TB, then we can consider how to take the 80T moker-vet share in smaller bites to move off the moker-vet pool.

Post new diagnostics after you have moved those other shares to the array.

Any update on that?

Hi Trurl,

On 2/23/2026 at 4:42 PM, trurl said:

Any update on that?

Yes, For some reason my mover didnt move it so i started doing it by hand with unbalanced. currently moving the chunks. i will update you when its done,

i expect the new 22tb disk today so i will install it inmeadly in the array as parity

On 2/19/2026 at 4:19 AM, Moker-Vet said:

I have set all shares (except moker-vet) that exist on moker-vet pool to move to array, this includes:

backup

downloads

media-landingzone

synology-nas

3 hours ago, Moker-Vet said:

For some reason my mover didnt move it

Post new diagnostics, those last ones you posted didn't show any of those shares configured to be moved.

moker-vet-diagnostics-20260226-0857.zip

Hi Trurl,

sadly the 22tb disk is not valid for a parity disk as it shows smart errors, i bought it second hand for a really cheap price (100€). i think we can use it as a temporal storage and later i could use it as a spare disk for transfer. but i dont feel comfortable to have it in the array for now. what do you think? it also doesnt let me do a smart test.

i am at the last 1 tb moving and then we can start moving the moker-vet pool, and moving the disks. i have the new 22tb as a unassigned device ready to be used as transfer storage.

On 2/25/2026 at 9:32 AM, trurl said:

those last ones you posted didn't show any of those shares configured to be moved.

And these latest diagnostics seem to show you renamed some of them, and I guess moved them manually since they aren't configured to be moved.

Only 2 shares still on moker-vet pool, the moker-vet share, not configured to be moved and won't fit anyway.

And this one which is configured to be moved but stiil has some files on mover-vet pool.

s----------s                      shareUseCache="yes"     # Share exists on moker-vet, disk3, disk4, disk5, disk9, disk11

Mover won't overwrite files, so possibly those files on moker-vet pool are duplicates of files on the array. You would have to examine that yourself.

7 hours ago, Moker-Vet said:

sadly the 22tb disk is not valid for a parity disk as it shows smart errors

7 hours ago, Moker-Vet said:

i have the new 22tb as a unassigned device ready to be used as transfer storage.

Is this the same disk that you said has SMART problems, or a different 22TB disk?

In any case, you can't assign a 22TB disk to the array except as parity since parity isn't that large.

We were planning to get a 22TB and some 20TB from moker-vet pool to reuse as array disks.

i have moved all files except moker-vet of the pool. im ready to proceed.

moker-vet-diagnostics-20260227-0941.zip

here is a new diagnostics just in case

7 hours ago, Moker-Vet said:

ready to proceed

There is an unassigned 22TB drive plugged in as sdx in those diagnostics, serial ZX2083X0. I assume this is the bad drive you referred to above. It is not suitable for any purpose.

# 1  Extended offline    Completed: read failure       90%       977         15738745840

This drive should be completely removed from your server, we won't be using it. Return it if you can.

Now we come to the part I have no personal experience in. Removing drives from btrfs raid5 pool.

Might be useful to review the thread starting here:

https://forums.unraid.net/topic/196809-varlog-is-getting-full/page/4/#findComment-1606177

We will be using this command line

btrfs dev del /dev/sdX1 /mnt/mover-vet

substituting the X in sdX1 for the actual letter designation of the drive we are removing.

Let's start with 22TB moker-vet (sdu), serial 00007NP2, which we will use as parity in the array after it has been removed from moker-vet pool.

So, the command would be

btrfs dev del /dev/sdu1 /mnt/mover-vet

But confirm that sdu is still the correct letter designation before you actually execute it. Sometimes those letters will change if things are plugged, unplugged, or just rebooted.

Be sure to run commands in a tmux session, I expect they will take a while.

When the command completes (prompt returns) post new diagnostics.

ok, i will read onto the ways on how to do this and once i have done it succesfully i will post the new diagnostics!