Missing "ProcessTrapsFromUnmanagedNodes" entry in Server Config

Started by Eisbaeeer, July 11, 2019, 09:53:43 AM

Previous topic - Next topic

Eisbaeeer

Hi
Im getting alarms from snmp traps of unmanaged nodes. After searching for solution, I found some threads about the server setting "ProcessTrapsFromUnmanagedNodes".
I´m running netxms 2.2.12.
This are the results for searching trap in server settings (see attached screenshot).
Any idea?
Kind regards Lars


Eisbaeeer

Ok, I tell you "from unmanaged node". This is not correct. The node is in managed mode, only the interface is in unmanaged mode.
I think, there is no setting in server-config, because this is by default, that no traps are processed in unmanaged mode, isn´t it?

So there is no way to disable snmp-traps from managed node in case of unmanaged interfaces?

Hope you guys understand my description ;-)

sincerly Lars

paul

No - pretty confused.

I have about a thousand devices unmanaged and they send traps and do not turn into alarms.

I have interfaces unmanaged but an interface is not a node - alarms register against a node.

Is the issues you have is that traps are being sent from an unmanaged interface on a node - or - is NetXMS monitoring an unmanaged interface and generating an alarm at the Node level?

I am pretty sure you cannot (currently) disable traps at an interface level as the traps will always aggregate up to the node - which makes sense as it is the node which generates the traps.

If you have created interfaces as nodes, then the fact they are unmanaged would come under the Node rule and would not trigger.

Excluding the fact that if a device wants to send your NetXMS server a trap, and you don't want to see it solely because it is coming down an unmanaged interface, you could probably filter this out using some scripting on the Event processing policy.

For me, I got annoyed with Interface Down alerts - so I disabled the policy - no more alerts. Either the other end of the interface is a node and I will get a Node down alert, or I have redundancy in which case the redundancy kicks in - and no action required. For those I really care about, I have a container limited policy and apply alarms to those only.

There is no setting as far as I know for what you want, but what you want can be achieved a number of ways as per above. I hope this helps.