Avoid false actions

[ittec]

Hi

recently I have noticed that in some cases, Zabbix send a alert message(smtp) when I restart some service like Apache. If the period of restart is higher than some secons, Zabbix agentd send warning to Zabbix Server and I get a message. But is not really a big deal, is only a restart

I don't know if there are some way to specificate a time before Zabbix send the message. I found some stuff about Time Period on Actions but I don't think is the point in my case.

Do you have some experience about it?

Thanks!

[tighep]

Sounds like you could use escalations to accomplish this. Make the steps like 2-0 and set whatever interval you'd want. Or set the steps to 2-2 if you only want one message. See if that will do what you need.

[ittec]

Thanks

Hi thigep

Unfortuneatly I have never read something about scalations, no experience. Im going to find some about scalations.

Ill try to explain once again my "problem" (is not a real problem, Im only trying to get more accuracy in actions). Example:

We have one host. With one Apache Web Server inside this host. We have then, one trigger to apache and one action.

If a crash of Apache starts, trigger and actions works good and send me a message. But, if you, personally, do a scheduled restart, Zabbix detects it too. And send another message, but this second is "wrong" because I do the restart myself. Was not an error.

This is the key!

Anyway, Im going to find something about scalations or whatever.

Thanks!

[ittec]

Question

I found this post on forums

http://www.zabbix.com/forum/showthread.php?t=11433

In one of catpures I noticed the version server is 1.6.1, mine is 1.6. Is possible that escalations were introduced on 1.6.1? Becasue I don't see the same in my server, the part of Actions Operations.

Thanks

[tighep]

I may have misunderstood your original question, let me try and understand better. It sounds like you've got some check for the apache service to be running. When apache crashes you get an alert from Zabbix, and that is the d esired behavior. When you restart apache you also get the alert, and that is not desired. Is that correct?

How often are you checking the apache service?

Can that time be increased to allow a restart without an alert?