Why is this trigger firing?

[rcollier]

Hello All,

I have a Zabbix item that checks the status of a database trigger. The Zabbix agent runs a script that outputs the status of the trigger. The value is then forwarded to the Zabbix server.

Here is a screenshot of the Zabbix Agent Item:


Here is a screenshot of the Zabbix Trigger for the item above


The idea here is simple. I only want to be alerted when the value is less than 1 for a period of 10 minutes. Right now, the trigger is firing whenever the value is less than one (regardless of how long it has been).

Is there something obvious here that I am not looking at? I have plenty of other triggers that use the min function along with a period of time. Why is this trigger firing when it shouldn't be?

Here are some additional screenshots to help:
Latest Data Showing the Trigger being less than one for less than 10 minutes.


Event Screenshot:


Any assistance would be appreciated.

Thanks

[ArtemK]

If I understand correctly, your trigger is ok, works as exptected, as soon as value is less than 1 in period of 10 minutes - trigger goes in problem state. I think you can define your trigger as max(10m)<1. In this case you will get trigger in problem state, only if max value of your item in period of 10 minutes is less than 1.

[kjetil_maland]

min(600)<1 translates to :

True If you in the last 600 seconds find a value that is less than one

What you want is

max(600) < 1

Which translates to
True If you in the last 600 seconds measure only values less than one


Mind you, now you have 10 minutes where things are bad.
You could do:

.count(10m,lt,1)>3

Which would rise the alert if you measure more than trhee values less than 1 the last ten minutes.

This of course will also fire after three conecutive less than one measurements.. but it will catch those instances where things are 'on and off'.

[karmen]

dudas en Triggers

Hola comunidad zabbix soy nueva en esta herramienta la cual conforme pasa
el tiempo le voy entendiendo ya que con la solución que ustedes va poniendo aqui...

Tengo una duda como hacer que un tigger que se ejecute solo tres veces por minuto cada intento para restablecer el servicio en este caso HTTP pero que no mande correo como alerta sino cuando el trigger ya se haya ejecutado esas tres veces y no se haya podido restablecer el servicio ahora si mande el correo de alerta que no se pudo estabilizar el servicio, ahora si en eso tres intentos o menos se estabilizo el servicio que mande una alerta que el servicio
se estabilizo