I have a btrfs that ran out of metadata space. Everything that matters has been copied off, but it's educational to try and recover it.

Now from when the btrfs is mounted R/W , a timer starts to a kernel panic. The kernel panic for the stack of "btrfs_async_reclaim_metadata_space" where it says it runs out of metadata space.

Now there is space data space and the partition it is on has been resized. But it can't resize the partition to get the extra space before it hits this panic. If it's mounted read only, it can't be resized.

It seams to me, if I could stop this "btrfs_async_reclaim_metadata_space" process happening, so it was just in a static state, I could resize the partition, to give it breathing space to balance and move some of that free data space to metadata free space.

However none of the mount options of sysfs controls seam to stop it.

The mount options I had hope in were skip_balance and noautodefrag. The sysfs control I had hope in was bg_reclaim_threshold.

Ideas appreciated. This seams like it should be recoverable.