I can still reproduce the problem after a rollback that snapper-cleanup service invocations fail. The output of `/usr/lib/snapper/systemd-helper --cleanup` for me is:
```running cleanup for 'root'.
running number cleanup for 'root'.
Deleting snapshot failed.
number cleanup for 'root' failed.
running timeline cleanup for 'root'.
running empty-pre-post cleanup for 'root'.
```
exit code is 1.
Same as in https://github.com/openSUSE/snapper/issues/159 in our case the problem was that there were volumes as particular btrfs subvolumes coming from docker or podman that prevented the deletion of snapshot id 1. We solved that by manually removing the nested subvolumes with `btrfs subvolume delete …` and then restart the snapper-cleanup service.