Hi everyone 

I had a cache pool of 2 X 1TB SSD in Raid 1 there was a docker issue so I ran out of space to solve the issue I added a 3rd 500GB SSD to the pool and I solved the docker issue . 

Now I`m trying to remove the 3rd SSD that I added but but with no luck 

 

here is what I tried 

Attempt  1

Unassign the 3rd slot of the cache pool

result : after starting the array (the pool is unmountable both 1TB SSD need to be formatted) 

Attempt  2

running the below command i got from Unraid Docs 

btrfs dev del /dev/mapper/sdX1 /mnt/cache 

Response : ERROR: not a block device: /dev/mapper/sdj1

 

Thanks In Advance 

Please post the diagnostics.

as requested 

trooper-diagnostics-20230913-1200.zip

Sep 13 00:51:15 Trooper emhttpd: shcmd (2013): /sbin/wipefs -a /dev/sdj
Sep 13 00:51:15 Trooper root: wipefs: error: /dev/sdj: probing initialization failed: Device or resource busy

 

Removed device failed to wipe because it was busy, reboot, don't start the array, post new diags and the output of

btrfs fi show

 

New digs attached 

 

output from command 

Label: none  uuid: ce49b453-e0bc-4321-9900-c7b565b0c2e9
        Total devices 3 FS bytes used 57.98GiB
        devid    1 size 931.51GiB used 90.03GiB path /dev/sdc1
        devid    2 size 953.87GiB used 91.03GiB path /dev/sdd1
        devid    3 size 465.76GiB used 21.00GiB path /dev/sdj1

Label: none  uuid: 664688eb-99b5-4f76-a6b8-6fe3461d349e
        Total devices 1 FS bytes used 351.45GiB
        devid    1 size 931.51GiB used 354.02GiB path /dev/sdi1

 

trooper-diagnostics-20230913-1251.zip

Start array with all 3 pool devices assigned as they are and post new diags to confirm pool profile.

trooper-diagnostics-20230913-1413.zip

Try removing the device again, stop array, unassign device you want to remove, start array, post new diags.

Yayyyy it worked 

I disabled both VM and docker first

then stopped the array unassigned the device started the array back up btrfs action started and all done within 5 Min

 

Thank You so Much 

trooper-diagnostics-20230913-1755.zip

The problem before was the disk busy error, though unclear what caused it, but usually a reboot fixes it.

It might be Docker and VM were turned on in the setting so unraid was protected the data in the cache pool 

because I tried to unassign without turning them off multiple times always getting the same result (the pool is unmountable both 1TB SSD need to be formatted)

 

anyhow I`m glad that it worked 

 

Thanks again

9 minutes ago, captpower said:

It might be Docker and VM were turned on in the setting so unraid was protected the data in the cache pool 

You need to stop the array to unassign a device, so those services must be stopped first, or the pool won't unmount, it was the device itself that was busy, not the pool, I've seen it before even in different scenarios, though still not clear the cause.