[Pkg-clamav-devel] Bug#775112: Bug#775112: systemd: repeatedly tries to start clamav

Andreas Cadhalpun andreas.cadhalpun at googlemail.com
Sun Jan 11 19:56:31 UTC 2015


Hi,

On 11.01.2015 20:17, Michael Biebl wrote:
> Am 11.01.2015 um 16:14 schrieb Russell Coker:
>> After a fresh install of a mail server running Jessie I get the above
>> repeatedly in my daemon.log until the filesystem becomes full.
>>
>> When I run "journalctl -u clamav-daemon" I just get the same as the above
>> with no additional information.  What should I do next to try and track this
>> down?

Please provide the output of:
sudo systemctl status clamav-daemon.service
sudo systemctl status clamav-daemon.socket
sudo systemctl status clamav-freshclam.service
ls -l /var/lib/clamav

> Looking at the file list of the clamav-daemon package [1], it seems to
> ship a clamav-daemon.socket unit. So it's most likely, that systemd
> starts clamav-daemon.service, because there is traffic on the socket.

Indeed.

> Depending on the clamav-daemon socket configuration [2], this will spawn
> a new service on each incoming request.

Only one clamav-daemon service is started at a time (Accept=false).

> I'm going to re-assign this bug report to clamav-daemon (not sure, if
> it's actually a bug), since its package maintainer is probably more able
> to help you.
>
> If this turns out to be an actual bug in systemd, please re-assign back.

My guess is that the clamav databases are not present.
The clamav-daemon.service contains:
ConditionPathExistsGlob=/var/lib/clamav/main.{c[vl]d,inc}
ConditionPathExistsGlob=/var/lib/clamav/daily.{c[vl]d,inc}

So systemd will not start clamav-daemon, if the databases are absent 
(and if it did, clamd would fail to start).
But these conditions should probably also be in the socket, so that the 
socket is not created, if no databases are present.

Anyway, freshclam should have downloaded the databases in the mean time, 
so it should work now.

Best regards,
Andreas



More information about the Pkg-clamav-devel mailing list