I'm building RAID5 out of 3 devices (3.7TB each). I run:
btrfs balance start -dconvert=raid5 -mconvert=raid5 /mnt/storage0
and it did process 3471 out of 3472 (sic!) chunks after which it failed with ENOSPC error.
/mnt/storage0 was previously a RAID0 volume composed of 2 devices (again, 3.7TB each, I then added thrird drive). I freed some space from the volume and started the process again. Unfortunately it'll take two days for it to complete and I'm wondering if there's better way of doing it? Is it possible to somehow resume
btrfs balance -mconvert -dconvert after it had failed due to ENOSPC?