What is be able to restart zabbix-server at midnights?

[HarryKalahan]

Hi All,

We've detected zabbix-server is restarting itself just at midnight [0:00].

We are monitoring the process with monitrc and it detects a change of PID of zabbix-server every night:

[CET Nov 5 06:25:01] info : 'system_insight' Monit reloaded
[CET Nov 6 00:00:13] error : 'zabbix_server' process PID changed from 10384 to 4822
[CET Nov 6 00:00:23] info : 'zabbix_server' process PID has not changed since last cycle
[CET Nov 6 06:25:01] info : Awakened by the SIGHUP signal
[CET Nov 6 06:25:01] info : Reinitializing monit - Control file '/etc/monit/monitrc'



I read zabbix.log and it appears the following messages:

16754:20141105:000007.499 Got signal [signal:15(SIGTERM),sender_pid:10378,sender_uid:0,r eason:0]. Exiting ...
16756:20141105:000007.499 Got signal [signal:15(SIGTERM),sender_pid:10378,sender_uid:0,r eason:0]. Exiting ...
16757:20141105:000007.500 Got signal [signal:15(SIGTERM),sender_pid:10378,sender_uid:0,r eason:0]. Exiting ...
16753:20141105:000007.501 Got signal [signal:15(SIGTERM),sender_pid:10378,sender_uid:0,r eason:0]. Exiting ...
16755:20141105:000007.502 Got signal [signal:15(SIGTERM),sender_pid:10378,sender_uid:0,r eason:0]. Exiting ...
16752:20141105:000007.502 Got signal [signal:15(SIGTERM),sender_pid:10378,sender_uid:0,r eason:0]. Exiting ...
16652:20141105:000007.503 One child process died (PID:16756,exitcode/signal:255). Exiting ...
16652:20141105:000007.610 syncing history data...
16652:20141105:000007.930 syncing history data done
16652:20141105:000007.930 syncing trends data...
16652:20141105:000007.939 syncing trends data done
16652:20141105:000007.939 Zabbix Server stopped. Zabbix 2.0.6 (revision 35158).
10384:20141105:000007.962 Starting Zabbix Server. Zabbix 2.0.6 (revision 35158).


I was checking several possibilities like database partition which it's executed at midnight by cron, but it doen't produce zabbix restart.

I was checking logrotate, but it's executed every day at 6:25.

Do you think what process in my Debian distro may produce zabbix-server restart?

I was searching on the net by "zabbix one child process died" but I coudn't find anything similar to my case. I've found a lot of links but none of them helped me.

Thank you for your help.

Best regards,

HarryKalahan.

[tchjts1]

I have no guess what is restarting it, but might I suggest upgrading your Zabbix server to at least 2.0.9? That is what I am currently running.

All you would need to do is compile the source and replace your existing binaries and restart the Zabbix server process.

I believe the earlier 2.0.x versions had some minor issues that made me move up to 2.0.9

[HarryKalahan]

Thank you for your answer, but this issue doesn't happen in all our customers, and we have customers with version 2.0.6 and 2.2.2, and it happens in both versions. I will continue investigating.

Any help would be appreciated.

Thank you very much!