Just noticed these errors in the log. sdf is one SSD out of 2 in my cache pool.   From other threads I think I need to run BTRFS scrub.   I have stopped all VMs and docker.   Before I did that I tried to take a backup of the cache drive to an unassigned disk, however some files failed with "cannot access source file"   I have daily snapshots of my VMs scripted and my VMs are in a subvolume.   Trying to copy one of the vm snapshots also resulted i