issue with PMG-Smtp filter

[killmasta93]

Hi,
I was wondering if someone could shed some light on the issue im having. running proxmox newest version
this is the error

root@mail:~# service pmg-smtp-filter status
● pmg-smtp-filter.service - Proxmox SMTP Filter Daemon
   Loaded: loaded (/lib/systemd/system/pmg-smtp-filter.service; enabled; vendor preset: enabled)
   Active: failed (Result: exit-code) since Thu 2019-02-21 05:02:06 -05; 6h ago
  Process: 1875 ExecStart=/usr/bin/pmg-smtp-filter (code=exited, status=0/SUCCESS)
 Main PID: 1905 (code=exited, status=111)

Feb 21 04:58:06 mail pmg-smtp-filter[1905]: starting database maintainance
Feb 21 04:58:06 mail pmg-smtp-filter[1905]: end database maintainance (6 ms)
Feb 21 05:00:06 mail pmg-smtp-filter[1905]: starting database maintainance
Feb 21 05:00:06 mail pmg-smtp-filter[1905]: end database maintainance (5 ms)
Feb 21 05:02:06 mail pmg-smtp-filter[1905]: starting database maintainance
Feb 21 05:02:06 mail systemd[1]: pmg-smtp-filter.service: Main process exited, code=exited, status=111/n/a
Feb 21 05:02:06 mail systemd[1]: pmg-smtp-filter.service: Killing process 1918 (pmg-smtp-filter) with signal SIGKILL.
Feb 21 05:02:06 mail systemd[1]: pmg-smtp-filter.service: Killing process 1919 (pmg-smtp-filter) with signal SIGKILL.
Feb 21 05:02:06 mail systemd[1]: pmg-smtp-filter.service: Unit entered failed state.
Feb 21 05:02:06 mail systemd[1]: pmg-smtp-filter.service: Failed with result 'exit-code'.

i was looking though the logs and also found this

Feb 21 01:00:26 mail pmgpolicy[28629]: 2019/02/21-01:00:26 Server closing!
Feb 21 01:00:26 mail pmgpolicy[28629]: 2019/02/21-01:00:26 Re-exec server during HUP
Feb 21 01:00:27 mail pmgpolicy[28629]: WARNING: Pid_file created by this same process. Doing nothing.
Feb 21 01:00:27 mail pmgpolicy[28629]: WARNING: Pid_file created by this same process. Doing nothing.
Feb 21 01:00:27 mail pmgpolicy[28629]: 2019/02/21-01:00:27 main (type Net::Server::PreForkSimple) starting! pid(28629)
Feb 21 01:00:27 mail pmgpolicy[28629]: Binding open file descriptors
Feb 21 01:00:27 mail pmgpolicy[28629]: Binding to TCP port 10022 on host 127.0.0.1 with IPv4
Feb 21 01:00:27 mail pmgpolicy[28629]: Reassociating file descriptor 5 with TCP on [127.0.0.1]:10022, using IPv4
Feb 21 01:00:27 mail pmgpolicy[28629]: Group Not Defined.  Defaulting to EGID '0'
Feb 21 01:00:27 mail pmgpolicy[28629]: User Not Defined.  Defaulting to EUID '0'
Feb 21 01:00:27 mail pmgpolicy[28629]: Setting up serialization via flock
Feb 21 01:00:27 mail pmgpolicy[28629]: Policy daemon (re)started
Feb 21 01:00:27 mail pmgpolicy[28629]: Beginning prefork (5 processes)
Feb 21 01:00:27 mail pmgpolicy[28629]: Starting "5" children
Feb 21 01:00:27 mail systemd[1]: Started Hourly Proxmox Mail Gateway activities.
Feb 21 01:01:04 mail postfix/anvil[793]: statistics: max connection rate 1/60s for (smtpd:185.234.217.221) at Feb 21 00:57:43
Feb 21 01:01:04 mail postfix/anvil[793]: statistics: max connection count 1 for (smtpd:185.234.217.221) at Feb 21 00:57:43
Feb 21 01:01:04 mail postfix/anvil[793]: statistics: max cache size 1 at Feb 21 00:57:43
Feb 21 01:01:43 mail pmg-smtp-filter[29302]: starting database maintainance
Feb 21 01:01:43 mail pmg-smtp-filter[29302]: end database maintainance (5 ms)
Feb 21 01:02:37 mail pmgpolicy[28629]: starting policy database maintainance (greylist, rbl)

Thank you

 

[Stoiko Ivanov]

Hi,

the pmg-smtp-filter error exit-code is a different one than in the original thread which you brought up (there it was 110, here 111) - I suppose the errors correspond to ETIMEOUT and ECONNREFUSED respectively see e.g. [0]

Does your system have enough memory - check `dmesg` and `journalctl` around the time pmg-smtp-filter parent process exits.

The second log seems ok from a quick look!

hope this helps!

[0] http://www-numi.fnal.gov/offline_software/srt_public_context/WebDocs/Errors/unix_system_errors.html

 

[killmasta93]

Hi,

the pmg-smtp-filter error exit-code is a different one than in the original thread which you brought up (there it was 110, here 111) - I suppose the errors correspond to ETIMEOUT and ECONNREFUSED respectively see e.g. [0]

Does your system have enough memory - check `dmesg` and `journalctl` around the time pmg-smtp-filter parent process exits.

The second log seems ok from a quick look!

hope this helps!

[0] http://www-numi.fnal.gov/offline_software/srt_public_context/WebDocs/Errors/unix_system_errors.html

Thanks for the reply, currently has 2 gigs of ram, whats odd this only happens when i implement the tutorial of the OṕenDKIM and the service always turns off around 4:50am very odd

 

[Stoiko Ivanov]

whats odd this only happens when i implement the tutorial of the OṕenDKIM and the service always turns off around 4:50am very odd

hmm - if it's always around the same time I'd take a look at cronjobs and systemd timers running around that time:
* '/etc/cron.d', '/etc/cron.daily', '/etc/cron.hourly', '/etc/cron.weekly',' /etc/cron.monthly', '/var/spool/cron/crontabs' - should be the places to look for cronjobs
* ` systemctl list-timers -all` - provides information about systemd timers

hope this helps!

apart from that 2g is afaik the minimum ram-setting (without extra software configured) - maybe just try to increase it and see if the problem goes away.

 

[killmasta93]

Thanks for the reply, didnt find nothing, nor adding to 4 gigs of ram helped either. i was guessing maybe creating a bash script but when i try it says service not found

when i ran didnt find the smtp pmg filter

root@mail:~# service --status-all
 [ + ]  clamav-daemon
 [ + ]  clamav-freshclam
 [ - ]  console-setup.sh
 [ + ]  cron
 [ + ]  dbus
 [ + ]  fail2ban
 [ + ]  fetchmail
 [ - ]  hwclock.sh
 [ - ]  keyboard-setup.sh
 [ + ]  kmod
 [ - ]  lvm2
 [ + ]  lvm2-lvmetad
 [ + ]  lvm2-lvmpolld
 [ + ]  networking
 [ - ]  nfs-common
 [ - ]  opendkim
 [ + ]  opendmarc
 [ + ]  postfix
 [ + ]  postgresql
 [ + ]  procps
 [ + ]  rpcbind
 [ + ]  rrdcached
 [ - ]  rsync
 [ + ]  rsyslog
 [ + ]  ssh
 [ - ]  sudo
 [ + ]  udev

 

[Stoiko Ivanov]

PMG uses systemd (like many linux distros) - try using `systemctl` instead of `service`

 

[killmasta93]

Thanks for the reply that did the trick im going to leave one day with every hour a cron job to restart it to see what happens