SNMP shows unknown state when strings are correct and SNMP validation works

I have recently upgraded from Orion 9.1 SP4 to SP5 Since then I am finding a problem with snmp managed devices. They are entered in with the correct strings and work fine until the devices looses contact for a short time. When my device returns, SNMP shows an unknown state ( this does not occur to all devices that loose contact its very random far as I can see). I can go into Orion Manager and SNMP validation works, I can even list resources. Once this occurs, the only way to fix it is to remove the device from Orion and re-add it. This looses all my history. I have tried to renumber the device, manage, unmanaged and re-adding the IP again. Only removal and readding seems to fix it. This is random across device platforms and snmp versions

Find more posts tagged with

orion_npm_9.1_sp5

SNMP

lose_contact

unknown

Accepted answers

All comments

male_pune4u

I have same issue, even i am not able to delete that nodes.

Also I have issue for 9.5 SP5 , I have newly added one node, however the device details showing is not of same device.

details are fetched from other existing node. not sure why NPM is not showing of same device.

please share if you have any info/resoluation for this issue.

bshopp

So why is the device loosing contact? Is it rebooting?

mgibson

Have you found a solution to this issue? I have the same issue except when I try to remove the node and re-add, I get and snmpv3 error encryption requested, but no short...

mgibson

Case 306217, I have replaced the solarwinds.net.snmp.dll files on all my pollers at supports direction. basically went from a v10.2.1 to v10.2.0.

It appears this still has not solved the issue.

Does anyone have any insight into this issue?

Thanks

epenney

We started noticing same behaviour since upgrading from 10.1.3 -> 10.3. We have multiple pollers, they are balanced, and appear healthy. I've ruled out its not a particular poller, but random interfaces. Definitely SNMP related. I've noticed that SNMP process appears to be hogging cpu on nodes being polled more so since upgrade. The node then appears to be undergoing SNMP denial behaviour and not responding, then orion reports 'unknown'.