False Positive on nodes - High response time over 200ms NPM

Hello!

I have seen several posts reporting false positives of high response times on network devices. I registered a ticket but I honestly didn't like the solution they recommended.

The problem only occurs on Node 1, I apply HA to Node 2 and the problem stops.

My environment is 2 virtual machines with 52% capacity polling, Solarwinds NPM 2024.1.1, 64 GB RAM, 14 cores CPUs 2.14 ghz, Intel Xeon, Windows Server 2022

Does anyone know what's happening? How should I start diagnosing to find the problem?

Sorry my bad english

Greetings!

Find more posts tagged with

false positive

High Response time

npm

Accepted answers

All comments

mesverrum

Are both servers on the same host and subnet, or are they in different parts of your data center in relation to the nodes being monitored?

If it was me i'd be running traceroute tests from both servers to the targets to see if they behave differently for any reason.

felipeaf88


Yes, both servers are on the same subnet. I am intrigued that this happens on server 1 only, server 2 does not have this problem (Virtual IP High Availability). Both servers have the same hardware

I will run tests that you recommend and I will tell you about them.

I currently have solarwinds services on node 2 (HA).





Regards!

stuartd

If all other things are equal, i.e. same subnet, same rules, etc then you need to be looking at the server and the connections from it.

Are the two VMs on the same host? If not, then check out the host. Do you have a.n.other host you can move the failing VM to?

adam.beedell

This is equal to "Poller 1 is all messed up" or perhaps "poller 1 is far away"

adam.beedell

@felipeaf88 to confirm, have solarwinds poller 1 monitor solarwinds poller 1, if the latency is still above ~0 then it means there's a delay is stuff leaving that box, consider it broken at that point.

Check TCP pot count, CPU/MEM, Element count, Reboot

vinay.by

I see that they have pitched in with great views, maybe it could a simple thing after all (I could be wrong) - make sure patching is completed on this server of yours and reboot it once. Post reboot check this again. Hope this helps.