You are not logged in.
This comes ups since systemd240 update for some services (auditd, glusterd, nmbd, smbd, winbind)
systemd[1]: /usr/lib/systemd/system/nmb.service:9: PIDFile= references path below legacy directory /var/run/, updating /var/run/nmbd.pid → /run/nmbd.pid; please update the unit file accordingly.nmbd example shown. So I update the units accordingly, restart the services or daemon-reload and all clear, then reappeared after today's systemd and samba updates. It's no harm left alone as systemd symlinks it anyway so I'm guessing its something the package maintainers haven't changed yet?
Last edited by WFV (2019-01-15 15:13:33)
∞ hard times make the strong, the strong make good times, good times make the weak, the weak make hard times ∞
Offline
The package maintainers are not responsible for changing it, since all these services come from the upstream sources.
I encourage you to submit upstream bugs to get these changed! Arch will not patch the upstream sources just to fix a warning.
Managing AUR repos The Right Way -- aurpublish (now a standalone tool)
Offline
Thanks Eli, I realized not a package maintainers issue after submit, apologies. Maybe not changed upstream for other reasons. Marking solved.
Last edited by WFV (2019-01-15 18:26:35)
∞ hard times make the strong, the strong make good times, good times make the weak, the weak make hard times ∞
Offline
It's a relatively new development by systemd! *Someone* needs to report it. ![]()
I'm reminded of the similar tmpfiles.d issues, for which I went and rebuilt lots of arch packages for the tmpfiles.d snippets which we authored, and submitted pull requests for some that were upstream. Sometimes not very successfully
https://github.com/fail2ban/fail2ban/pull/2174
Managing AUR repos The Right Way -- aurpublish (now a standalone tool)
Offline
Not sure what or where to report - have no experience there. There are a lot of unit files in /usr/lib/systemd/system but only those mentioned, also teamviewerd, posted warnings. I did as mentioned, all good, also had a reboot before the next update and warnings gone everything started normal with PID's from /run instead of /var/run. Then reappeared (including auditd which didn't update) after glusterfs, samba, and systemd update. I just reinstalled systemd and its counterparts (lib, lib32, sysvcompat) in a virtual machine to see it the warnings reappear and they don't. Can try reinstall the other packages again and see what happens. Should I unsolve this thread?
EDIT: reinstalling packages resets their pidfile to /var/run (samba example) as expected, possibly other package updates that affect Unit Files might also report the systemd warning.
Some more related material. Being /var/run already linked to /run seems the warning maybe a non-issue.
Last edited by WFV (2019-01-17 00:07:17)
∞ hard times make the strong, the strong make good times, good times make the weak, the weak make hard times ∞
Offline
Where do we update these unit files again?
Things are never what they seem...
Offline