We discovered that none of Cisco network devices from any dedicated Additional pollers (APE) are not polling data. In each DC dedicated HA APEs collecting data.
We got some data sporadically; eg randomly (as per graphs) each x weeks or month, this short poll (data collection) is happening when there is triggered some failover on APE or main orion server.
This seems never worked correctly as it was very hard to find out because: all sensors act as green with OK values as last data (old days, weeks, months) are HANGOUT,-->simply shows last poll data.
When I disable and enable sensor on any devices or any sensor (over 100 devices and 7000 sensors) I will get to the point that last value will be: UNKNOWN and grey out icon of the sensors.
We have open support case for now for 3 months with no solid progress. hence I am asking here, we tried with no progress:
- switching between 2 preferred MIB either globally or per device
- upgrade to latest Solarwinds 2020.2.6
- failover over to other main orion and APEs (active/passive change)
- on APEs and main orion servers (all in HA):
a) stop the collector service on the affected APE
b.) delete C:\ProgramData\SolarWinds\Collector\Queues\solarwinds.node.hardwarehealth.snmp.db
d.) Restart the collector service. This will recreate the .db file deleted in step 3
Any hint or what to try is appreciated. As we are missing basic feature of HW sensor asset monitoring (fan, power, rx/tx power on SFP, etc).
Just found this tool: https://support.solarwinds.com/SuccessCenter/s/article/Orion-service-manager-does-not-stop-Collector-Data-Processor?language=en_US
and will ask assigned tech to support case what we can do with that 
Thank you.
