zabbix stop sending email via postfix

[[email protected]]

Hi there

Our zabbix server has stopped sending emails.

I researched all possible posts in your forum but had no luck with it.

We use postfix as a SMTP server on the same box (localhost) and I can send email via command line fine (as a user which runs zabbix server).

I checked server the logs (DEBUG mode) but nothing there. We don't have zabbix_suckerd.log file available and server verison is zabbix-1.8.3rc3.

Emails worked fine but we realized than we haven't got an email alert for a while..

Please help.

Regards,
Rad

[makini]

Similar behavior...

We have version 1.8.4 that also stops sending email alerts after some time.
This behavior did not appear before, then the server process started crashing completely, with debug log and etc.

We suspected that an ipmi libs problem caused the crashes - the server was recompiled with ipmi support several days before that, but recompiling it back without the ipmi didn't help. It still stops sending alerts after some time and the server process keeps crashing on start (when restarted). The emails stop getting sent even before the server crashes, when it seemingly still runs normally (no errors in log).
Highly problematic behavior - it's our main production monitoring system...

Attached is a crash debug log (objdump of the binary is also available upon request - it's 17MB in size, can't be attached here). There's also a bug I opened about this on Zabbix Support.

[makini]

Our issue was resolved...

Though the solution wasn't highly effective and is more like a work-around...
There was a problem with Escalator process of zabbix_server, here's the relevant part of the crash log:

9641:20110414:174107.219 Escalation cancelled: Trigger 'Timezone Verification - {HOSTNAME}' disabled.
9641:20110414:174107.336 Escalation cancelled: Host 'pdc' disabled.
9641:20110414:174107.451 Got signal [signal:11(SIGSEGV),reason:1,refaddrnil)]. Crashing ...
...
9202:20110414:174107.464 One child process died (PID:9641,exitcode/signal:255). Exiting ...

The solution was to clean up the "escalations" table of the DB (it was also checked for corruption - didn't help), disable the escalations related Action(s) and start the server again.
Closed the original bug and opened a more relevant one.

[[email protected]]

Version

Hi there

and thanks for quick response.

I think I'll update zabbix version and re-install it.
I checked the "escalations" table and it was empty so this did not help.
Also I restarted zabbix server as well as the whole server with no luck. Using postfix from command line works fine but zabbix is not having it and no change with log files.

It's a bit manic so I'll have to find a slot and re-do zabbix server later on and I'll keep you updated.

Cheers,
Rad