problem with false triggers unreachable for 5 minutes

[Magician]

hi,

im having a problem from 3am until 5am everyday of host unreachable for 5 minutes .

when im looking in the zabbix history , the agent is up and it seems ok , even so in the agents themselves between hour 3-5 am it seems like it sends a problem with the status of up , here is the example :

Trigger: Zabbix agent on ********* is unreachable for 5 minutes Trigger status: PROBLEM Trigger severity: High Trigger URL:

Item values:

but as you can see here the event it self is reporting up -1. 10105 (*******:agent.ping): Up (1)

it is zabbix 2.0.3 and it happens just between 3am a-5am i checked with ping from another computer and all server are up and dont fall , what can you recommend to solve this problem ?

[joshuamcdo]

Re: Agent status

I had the same issues with hosts in the cloud until I upgraded to 2.0.9. I even tried jacking up the wait times, nothing helped until I went to 2.0.9.

J

[Magician]

thanks , but meanwhile i increased the max connections to 512 and the pollers to 1000 - this morning it didnt repeat itself , so ill give it a little time see how its working , if its repeating ill upgrade to 2.0.9 or 2.2

[trikke76]

when u see this kind of behaviour first thing u need to do is check the zabbix queue

probably there is a delay from some items ipmi, zabbix agent, .....

then u have an idea where your problem is.
increasing the pollers can help then in some cases.