Delta
24-12-2007, 10:18
I have a fairly new installation of Zabbix 1.4.4 (upgraded from 1.4.3, which I hoped would solve my problem below, which it didn't) on FreeBSD 6.2 that monitors correctly but unfortunately seems to crash when an action is performed.
The situation is as follows: about 15 servers are monitored. I can produce all graphs and when a host or service is down it is correctly displayed as down. Now I have created an action that sends me an email when one of the hosts is down. When I activate the action and when it sees that the host is down, zabbix_server crashes.
When I restart zabbix_server it crashes again after a few seconds. When I disable the action (or undo whatever is triggering the action), all is fine and zabbix_server will continue running.
I added the level 3 and level 4 debug output (I took out quite a few rows in the level 4 output that started with 'In delete_history' because otherwise it would be too large to post; even zipped) from zabbix_server.log. The zabbix_client processes on the monitored hosts are not affected and keep running without any errors.
In this case there is an action that triggers on Host availability of the host EU-DBTEST. As you can see at the bottom of the log, it correctly notes that the host is down and then crashes.
This is the message in /var/log/messages from the level 3 debug:
Dec 24 09:28:02 zabbix kernel: pid 59829 (zabbix_server), uid 1003: exited on signal 11
This is the message in /var/log/messages from the level 4 debug:
Dec 24 09:30:58 zabbix kernel: pid 59879 (zabbix_server), uid 1003: exited on signal 11
uid 1003 is the zabbix account
Does anybody have any idea? This is of course a show stopper, since Zabbix is supposed to notify me instead of crashing and thus stop monitoring at all.
The situation is as follows: about 15 servers are monitored. I can produce all graphs and when a host or service is down it is correctly displayed as down. Now I have created an action that sends me an email when one of the hosts is down. When I activate the action and when it sees that the host is down, zabbix_server crashes.
When I restart zabbix_server it crashes again after a few seconds. When I disable the action (or undo whatever is triggering the action), all is fine and zabbix_server will continue running.
I added the level 3 and level 4 debug output (I took out quite a few rows in the level 4 output that started with 'In delete_history' because otherwise it would be too large to post; even zipped) from zabbix_server.log. The zabbix_client processes on the monitored hosts are not affected and keep running without any errors.
In this case there is an action that triggers on Host availability of the host EU-DBTEST. As you can see at the bottom of the log, it correctly notes that the host is down and then crashes.
This is the message in /var/log/messages from the level 3 debug:
Dec 24 09:28:02 zabbix kernel: pid 59829 (zabbix_server), uid 1003: exited on signal 11
This is the message in /var/log/messages from the level 4 debug:
Dec 24 09:30:58 zabbix kernel: pid 59879 (zabbix_server), uid 1003: exited on signal 11
uid 1003 is the zabbix account
Does anybody have any idea? This is of course a show stopper, since Zabbix is supposed to notify me instead of crashing and thus stop monitoring at all.