Network Management

How to poll Down nodes in NPM

This article provides brief information on polling nodes that are down.

First published date

10/24/2018 5:19 PM

Last published date

5/29/2019 7:58 PM

Overview

This article provides brief information on polling down nodes.

Product section

Network Performance Monitor

Resolution

When Orion does not get a response from a node, we cannot immediately assume the node is down, as there are a number of other reasons why the node may not be responding other than it is being down. We want to be sure the node is down before we mark it as such. How do we do this: 

When a node goes down it may not go down for long, depending on your polling settings it may miss the node going down in which case the alert will not fire at all. 

Say you are polling every 120 seconds - which is the default polling, If a node goes down at the very start of the 120 seconds it has to wait 120 seconds before it gets polled again. When it gets polled again and does not respond Orion puts the node into warning status and then goes into Fast Polling where it polls the node every 10 seconds for 120 seconds. If the node responds then it will put the node back into an up state.

Orion is using ping (ICMP) or SNMP to get a response of the device. Make sure that these devices are reachable on your Orion server and no network connectivity issues so that you can avoid this kind of behavior.

Check:

  1. RDP to the server from which the node is being polled.

  2. Ping the node IP from this server:Ping the node IP from this server

  3. If the node does not reply, the network admin must investigate why the node is not responding to ICMP from the server it is being polled from.