[SOLVED] Proxmox mail gateway Clamav error

[jmaitra]

I successfully setup PMG on a Debian 10 and got it running so far.

First, clamav-freshclam failed so I removed apparmor service (as explained in the forum) and rebooted the machine. Now clamav-freshclam works fine but the clamav-daemon stops after about 30 seconds with the following error message

Oct 04 11:23:01 mars.maitra.biz systemd[1]: Starting Clam AntiVirus userspace daemon...
Oct 04 11:23:01 mars.maitra.biz systemd[1]: Started Clam AntiVirus userspace daemon.
Oct 04 11:23:01 mars.maitra.biz clamd[830]: Received 0 file descriptor(s) from systemd.
Oct 04 11:23:01 mars.maitra.biz clamd[830]: clamd daemon 0.102.4 (OS: linux-gnu, ARCH: x86_64, CPU: x86_64)
Oct 04 11:23:01 mars.maitra.biz clamd[830]: Running as user clamav (UID 109, GID 116)
Oct 04 11:23:01 mars.maitra.biz clamd[830]: Log file size limited to 4294967295 bytes.
Oct 04 11:23:01 mars.maitra.biz clamd[830]: Reading databases from /var/lib/clamav/
Oct 04 11:23:01 mars.maitra.biz clamd[830]: Not loading PUA signatures.
Oct 04 11:23:01 mars.maitra.biz clamd[830]: Bytecode: Security mode set to "TrustSigned".
Oct 04 11:23:45 mars.maitra.biz systemd[1]: clamav-daemon.service: Main process exited, code=killed, status=9/KILL
Oct 04 11:23:45 mars.maitra.biz systemd[1]: clamav-daemon.service: Failed with result 'signal'.
Oct 04 11:32:45 mars.maitra.biz systemd[1]: Starting Clam AntiVirus userspace daemon...
Oct 04 11:32:45 mars.maitra.biz mkdir[5222]: /bin/mkdir: cannot create directory â/run/clamavâ: File exists
Oct 04 11:32:45 mars.maitra.biz systemd[1]: Started Clam AntiVirus userspace daemon.
Oct 04 11:33:19 mars.maitra.biz systemd[1]: clamav-daemon.service: Main process exited, code=killed, status=9/KILL
Oct 04 11:33:19 mars.maitra.biz systemd[1]: clamav-daemon.service: Failed with result 'signal'.

Reinstall apparmor didn't fix the problem.

Any help would be greatly appreciated. Many thanks in advance
Jens

 

[Stoiko Ivanov]

does the system have enough memory?
please post the output of:
* `dmesg`
* the content of the logfiles in '/var/log/clamav'

 

[jmaitra]

Thanks for the fast reply. It seems that you are right - there is an out "of memory-error".


[81472.533565] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=/,mems_allowed=0,global_oom,task_memcg=/system.slice/clamav-daemon.service,task=clamd,pid=15850,uid=109

[81472.533583] Out of memory: Killed process 15850 (clamd) total-vm:1363408kB, anon-rss:1271548kB, file-rss:0kB, shmem-rss:0kB, UID:109 pgtables:2624kB oom_score_adj:0

[81472.688651] oom_reaper: reaped process 15850 (clamd), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB


But I only use about 40% of the available memory (see attached snapshot).

 

[Stoiko Ivanov]

How much memory does the PMG installation have assigned?

4g should work fine in the default setup

 

[jmaitra]

according to your documentation

https://www.proxmox.com/de/proxmox-mail-gateway/systemanforderungen

the machine has 2 only GB RAM - it's only a test environment

 

[Stoiko Ivanov]

* anything interesting in the clamav logs during the time when the OOM kill happened?
* if you can easily reproduce the issue it would be great if you could keep an eye on `htop` and see which processes eat up the memory.

Thanks

 

[jmaitra]

Stoiko,

thank you for your response. The issue is easily reproducable by starting clamav-daemon

# systemctl start clamav-daemon

analysis results with "htop":
clamd - process uses consumes all cpu-ressources
clamd - process uses consumes all memory ressources
after 20 seconds it will be killed by OOM

CLAMAV logs in /var/log/clamav remain empty


Best regards
Jens

 

[Stoiko Ivanov]

thank you for your response. The issue is easily reproducable by starting clamav-daemon

hmm - do you have PMG running inside a VM or a container? (tried reproducing the issue here with a debian container having 2G and the issue does not occur)
* do you have any additional signatures (or other config modifications for clamAV)?

Thanks!

 

[jmaitra]

I installed based on a debian 10 VM (1 vcpu, 2 GB RAM, 40GB SSD) i ordered at NETCUP GmbH

First I changed the hostname, afterwards, i changed /etc/apt/sources.list (due to test-evironment without subscription) and then

apt install proxmox-mailgateway

That's it (finally i only removed apparmor for getting clamav-freshclam started)

 

[Stoiko Ivanov]

Ok - managed to reproduce this here - seems clamav has gotten a bit more memory-hungry since we last updated the docs/specs (will send a patch)

The setup works with 2GB ram, if you have some swap-space available (1G should be enough)
also 3GB should be enough to get clamav started

I hope this helps!

 

[jmaitra]

I added 2GB of swapspace and configured clamav-daemon to a maximum of 30%cpu and 384mBytes of RAM

• vi etc/systemd/system/clamav-daemon.service.d/extend.conf
  
  IOSchedulingPriority = 7
  CPUSchedulingPolicy = 5
  MemoryLimit=384M
  CPUQuota=30%
  Nice = 19

all services are running so far

Thank you for your great support !

 

[Stoiko Ivanov]

all services are running so far

Glad that worked

Please mark the thread as SOLVED - this helps others who also run into the issue
Thanks!