Hello...
We monitor F5 devices in SolarWinds Orion NPM for memory, CPU and a number of other statisticsab
. As F5 devices do not give meaningful results for the standard CPU and memory OIDs we use a number of Universal Device Pollers (UnDPs) to gather this information.
The majority of the nodes we monitor in NPM get queried using the standard the standard CPU and memory OIDs. Sometimes the the SNMP process for a node fails or there is an issue with SNMP traffic between the poller and the node for example a firewall no longer passes the SNMP traffic. In these situations SolarWinds changes the status of the node from 'up'. This is good.
The issue we have is that if we have an issue with the SNMP process or the SNMP traffic associated with an F5 node we are not getting the same notification that there is an issue. We have found an F5 node that is no longer responding to SNMP but the status of the node is still 'up'. This is not good.
To ensure we are alerted when an F5 device is no longer responding to queries, I want NPM to raise an alert for all of the F5 nodes that have the UnDP called sysStatTmTotalCycles where it does not return a value.
I tried to set up an advanced alert with these Trigger Conditions:
Poller Name is equalto sysStatTmTotalCycles
Status is empty
This did not create an alert for the F5 device that is no longer responding to SNMP.
I tried to set up another advanced alert with these Trigger Conditions:
Poller Name is equalto sysStatTmTotalCycles
Numeric Status is empty
This also did not create an alert for the F5 device that is no longer responding to SNMP.
Does anyone out there know how to raise an alert in the situation where no value is returned by a Universal Device Poller (UnDP)? Any help gratefuly accepted.
Cheers,
... Simon Evans
