Quite often I cannot overlay packages using rpm-ostree. After the reboot I run:
$ rpm-ostree status
State: idle
Warning: failed to finalize previous deployment
ostree-finalize-staged.service: Failed with result 'timeout'.
check `journalctl -b -1 -u ostree-finalize-staged.service`
And indeed it did not finish - during the reboot it hung on the “OSTree Finalize Staged Deployment” service until it reached the timeout.
Inspecting the journal, I understand the os-prober scans all the partitions and more importantly all the btrfs subvolumes - this takes too long on my system and therefore breaks.
this is the end of journalctl -b -1 -u ostree-finalize-staged.service
:
Oct 03 23:46:39 root[15229]: os-prober: debug: btrfs volume uuid=UUID partition=/dev/sdc
Oct 03 23:46:39 root[15229]: os-prober: debug: running /usr/libexec/os-probes/50mounted-tests on btrfs /dev/sdc
Oct 03 23:46:39 root[15229]: 50mounted-tests: debug: begin btrfs processing for UUID
Oct 03 23:51:01 root[15229]: 50mounted-tests: debug: btrfs volume UUID mounted
Oct 03 23:51:33 systemd[1]: ostree-finalize-staged.service: Stopping timed out. Terminating.
Oct 03 23:51:33 systemd[1]: ostree-finalize-staged.service: Control process exited, code=killed, status=15/TERM
Oct 03 23:52:01 systemd[1]: ostree-finalize-staged.service: Failed with result 'timeout'.
Oct 03 23:52:01 systemd[1]: Stopped OSTree Finalize Staged Deployment.
-
Please how can I blacklist partitions so that os-prober would not mount and scan them?
-
And out of curiosity: If I disabled the os-prober (how would I do that please?) would it keep the entries for other OSs which it had already created in the past, or are these entries always recreated anew from scratch?