Discovery and nodata trigger...

[QwErTy_LoGiC]

Hello everyone,

I have a situation and I am wondering if this is normal.

I have a discovery setup to scan a subnet with both an icmp check and an snmp request. This part works perfectly. The scan is configured to every 4 hours.

I then have an action which checks if the received value is like Cisco and if the status is up and the uptime greater than 10 minutes. If all of these are true, the device is added to a group and then a template linked to it. This works fine to.

The problem is, in the template linked, there are two items, again an icmp ping check and an snmp request. The ping checks the responsiveness, then snmp request checks for a valid snmp configuration. Both items have triggers, and the SNMP trigger depends on the ping trigger. The ping trigger check that the last two pings were successful and the snmp trigger checks that a valid snmp response came through in the last 3 minutes. both items are set at 60 sec intervals.

The thing is, the SNMP trigger goes off when a device is discovered through this process...

Am I missing something in the process because I do not get why it goes off...

Any ideas?

Thanks!

[QwErTy_LoGiC]

Something more...

Hello all,

I have noticed something more, even though my agents on have an agent.ping active check, and that those checks are successful, in the host configuration screen, the zabbix icon stays grayed out. The agent.ping item is checked green for the host, the trigger is working, I can see the latest data for the agent.ping but it is as though the zabbix server is unable to calculate the availability status of the host...

I wonder if this is related to my previously mentioned issue...

Any ideas?

[QwErTy_LoGiC]

One down, one to go

Ok,

I changed the item type to Zabbix passive check, instead of Zabbis active check and everything came back up.

There is still the matter of assigning an item with a nodata trigger and it goes off as soon as it is assigned, only to go back to normal after a few moments...

This is annoying, if I assign a template to 300 devices, I'm gonna get 300 alerts and 300 recoveries (if these are configured) only because it determines it did not have any previous data (duh!) on the said item which was just assigned.