Time out on host : no triggers!

[Justleen]

Hi all,

We are changing our snmp community string at the moment, and I noticed something potentially dangerous (for us anyway).
We have 5 switches monitored with 1 template.
I have changed the community string on 1 switch, and changed the template community string.

This (obviously) works, on the switch that I changed.
The thing is, I was keeping an eye on the dashboard, waiting for 4 switches to throw a warning of some kind, but that never came. When checking the config for the 4 switches, I now see the tests disabled, with an error "timeout while connecting". Again, this is working as designed, but I would really would like to know why this doens't give any error or warning on the dash. As the situation is now, it means that a switch may die on me, and I'd never know.

Shoudl I create a trigger for cases like this myself? Are there any build-ins to catch this?

Template used is the Template_HP_Procurve which comes with the default install.


Thanks in advance,

Leen

[Calimero]

It won't give any error on the dashboard as you didn't tell zabbix to do so.

You should write a trigger on one of your SNMP items with the .nodata() function.

[troffasky]

Alternatively you could add a check on the SSH port or ICMP ping.

[Justleen]

It won't give any error on the dashboard as you didn't tell zabbix to do so.

You should write a trigger on one of your SNMP items with the .nodata() function.

I shouldnt have to tell zabbix that? I have other triggers (not actions!) which go onto the dashboard, without me explicitly telling zabbix so.

So now a whole template, with a bunch of triggers (that work), is unreachable and I get no errors or warnings..

Feel counter intuitive to me..

Ill go and create some nodata actions..