Comment 3 for bug 1804060

Revision history for this message
Christian Ehrhardt  (paelzer) wrote :

Hi Oliver,
I agree with Bernd on https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=856429#10
With sysV init being even "less meant to be considered" on Ubuntu than on Debian (and even there it is not more a big thing).

If:
$ journalctl -u vgauth.service
gives you the right output then all is fine.

I'd not want to touch sysV files these days anymore - nobody uses and nobody should use them anymore.

For systemd services are supposed to not daemonize to be more under control by systemd.
So adding -s which is defined as:
   printf("\t-s\tRun in daemon mode.\n");
is wrong IMHO.

I see and agree the argument of vmtoolsd having a local log file and that consistency would be nice.
But IMHO a solution to that would be more towards making vmtoolsd to log to journal one or another way. Is vmtoolsd currently logging.

If you or people "need" an extra log file I think something like an rsyslog rule in /etc/rsyslog.d/ might be appropriate.
One could map event sources to a file they are logged.
But since that is double-logging that usually needs a good reason.

Are there complains that the logs through journalctl are not accessible that would make this wanted?

Marking incomplete until clarified in a discussion if/what we would want to change.