Bug#844178: Better systemd integration

[Michael Biebl]

Hi Andreas

Am 08.02.2017 um 19:49 schrieb Andreas Metzler:

> It depends on what the majority wins at what loss for the minority. And
> I just do not know whether the benefits outweigh to cost.
> 
> I kind of buy your argument about "Does pidfile need to be configurable"
> and would be willing to move this to unsupported if there is some gain
> for the pain of breaking some systems.

The benefits are, that by specifying the pidfile, systemd is able to
properly track the daemon life cycle.
Atm, if you run service exim4 start, then kill the daemon, systemd still
thinks the service is running. This is highly confusing for users.

The problem is, systemd can't know without additional meta data, if a
service is supposed to start a long running process or not.
The # pidfile: comment (or a drop-in snippet for that matter) is such a
hint.

> What about the fact that the current setup allows to start two daemons
> from the init script (queue-runner and listener)? What changes when the
> pidfile setting is added, pointing to the listening daemon?

The pidfile points systemd at the "main" process which is considered
crucial for the service to be up.
If that PID dies, systemd will stop all processes in the cgroup.

Michael


-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <http://lists.alioth.debian.org/pipermail/pkg-exim4-maintainers/attachments/20170208/604fbb36/attachment.sig>