action #125132
closed[alert] logrotate failed on OSD
0%
Description
from journalctl:
Feb 15 00:00:07 openqa logrotate[12569]: logrotate does not support parallel execution on the same set of logfiles.
Feb 15 00:00:07 openqa logrotate[12569]: error: state file /var/lib/misc/logrotate.status is already locked
Feb 15 00:00:00 openqa systemd[1]: Starting Rotate log files...
Updated by okurz over 1 year ago
- Related to action #124412: [alert] logrotate services failed on openqa-piworker.qa.suse.de and OSD size:M added
Updated by mkittler over 1 year ago
- Status changed from New to Feedback
I haven't received a mail about failing systemd services today anymore. I suppose this ticket has been created for a mail from yesterday.
This mentioned log is even from 15 Feb. So it had happened before my changes for #124412 were deployed. The most recent encounter is:
sudo journalctl --since '1 day ago' -fu logrotate.service -u logrotate-openqa.service
…
Feb 27 15:00:00 openqa logrotate[10366]: uncompress_prog is now /usr/bin/xzdec
Feb 27 15:00:00 openqa logrotate[10366]: error: state file /var/lib/misc/logrotate.status is already locked
Feb 27 15:00:00 openqa logrotate[10366]: logrotate does not support parallel execution on the same set of logfiles.
Feb 27 15:00:00 openqa systemd[1]: logrotate-openqa.service: Main process exited, code=exited, status=3/NOTIMPLEMENTED
Feb 27 15:00:00 openqa systemd[1]: logrotate-openqa.service: Failed with result 'exit-code'.
Feb 27 15:00:00 openqa systemd[1]: Failed to start Rotate openQA log files.
Feb 27 15:48:25 openqa logrotate[29411]: error: destination /var/log/rsyncd.log-20230227.xz already exists, skipping rotation
Feb 27 15:48:25 openqa logrotate[29411]: error: destination /var/log/zypper.log-20230227.xz already exists, skipping rotation
Feb 27 15:48:25 openqa systemd[1]: logrotate.service: Deactivated successfully.
Feb 27 15:48:25 openqa systemd[1]: Finished Rotate log files.
…
and also that was (2 hours) before the MR has been merged.
I suppose we can keep this ticket open (in addition to #124412) as it is specific to the OSD problem only (although it is basically duplicating #124412).
Looks like my MR has still one mistake in it:
Feb 27 18:00:00 openqa bash[19337]: ++ /usr/bin/systemctl --user is-active logrotate.service
Feb 27 18:00:01 openqa bash[19340]: Failed to connect to bus: $DBUS_SESSION_BUS_ADDRESS and $XDG_RUNTIME_DIR not defined (consider using --machine=<user>@.host --user to connect to bus of other user)
Feb 27 18:00:01 openqa bash[19322]: + is_active=
Feb 27 18:00:01 openqa bash[19322]: + [[ '' == active ]]
Feb 27 18:00:01 openqa bash[19322]: + [[ '' == activating ]]
Feb 27 18:00:01 openqa bash[19322]: + exit 0
The --user
flag shouldn't have been used. I've created a MR to fix this (https://gitlab.suse.de/openqa/salt-states-openqa/-/merge_requests/800) and have also applied the change manually on OSD.
Updated by okurz over 1 year ago
- Priority changed from Urgent to Normal
https://gitlab.suse.de/openqa/salt-states-openqa/-/merge_requests/800 merged and effective, back to "Normal".
Updated by okurz over 1 year ago
- Status changed from Feedback to Resolved
We will know if the problem reappears from alerting.