Zabbix servers go down after n days or hours

[markusz]

Hi

I've compiled Zabbix server on three different machines (one Debian, two Redhat (Centos)).

On every system the server dies spontaneously after some days or even hours. I don't know why, I have no idea why this happens.

Is there some mechanism that causes the server to stop (and should eventually start it again)?

Starting the server manually via init scripts works - I'm just sick to do this every two days.

[untergeek]

I've had this problem sporadically also, but mine is usually due to database connectivity issues.

I've simply implemented SEC http://kodu.neti.ee/~risto/sec/ and created rules to tail the zabbix_server.log file to see if it quit and to send me notification when db issues arise. The rule I created auto-restarts on detected fail/shutdown.

I've also used SEC to run a pseudo-cron job that runs a script every 5 minutes to check to see if zabbix_server is running, and if not, start it again.

My problems have been much less irritating. I don't have restarts every two days (mine could be anywhere from a few hours to a few weeks, with weeks being the far more common), but when the zabbix server goes down, I get email notification and an attempted auto-restart. It's been nicer.