Created attachment 852174 [details]
dmesg, SysRq-L, perf data, lsblk -f, snapper list

My CPU fan had a long high-intensity exercise this morning. The root cause
turned out to be a kworker thread doing "btrfs_preempt_reclaim_metadata_space".
The offending thread identification is "kworker/u16:11".

I've collected some data bits (dmesg, SysRq-L, perf data) during the event. It
is unclear to me, however, what triggered this long-running FS maintenance and
how I could stop it in case I needed to - i.e. if the laptop ran battery or if
I sat in a quiet room where the CPU fan disturbance would be unacceptable.