Problem with hardware sensor alert...

Forgot to add that on the alert it is referring to a power sensor.

Vote on solving this issue! 

We commonly see this in our environment for 10g transcievers and on Catalyst 65xx transcievers. THEORETICALLY, it is alerting correctly. When you dive into the trans statistics you will see where it will show a -40db loss or a amp or Mamp value that is outside of the switches alerting thresholds even when the port is not in use, from the SFP being fully slotted. Now, Cisco knows that when a port is admin shut, the box isn't going to scream at you for the values being outside of thresholds, because well we aren't using those transceivers; Solarwinds on the other hand when it polls, it sees those values, says "hey this is out of my threshold, ALERT", falling into that critical hardware out of box alerting. Unfortunately, with it being an out of box alert, i don't believe you'll be able to modify it until Solarwinds can address this internally with a hotfix or patch. It would be nice to see them add the logic in the alert to detect if interface is admin down to ignore transceiver alerts.

We have the same problem.  By the way, in our case, we do not monitor interfaces that are in ADMIN DOWN state.  You would think that SolarWinds would be smart enough to ignore interfaces that are in ADMIN DOWN status.

Also, I called SolarWinds tech support and they said I'd need to manually disable the hardware sensors for the ADMIN DOWN interfaces.  That's no fun!!!   

Anybody know if there has been any movement on this within solarwinds?  Seems to still be an issue after all these years.  

Hi Everybody! I just wanted to take a crack at this detail that we see sometimes in polling. We do have two options to poll hardware with, and the default is the Entity-sensor MIB. This table is a huge table with a lot of details, not everything has a serial#, not everything has a status/metric. 

I used to poll this table manually with some UnDP's.  To take a different approach to your hardware edit those nodes and change to the CISCO-ENVMON-MIB, instead of the Entity MIB.

You should see switch details, fans, Power supplies and temps for the switches/line cards instead of the granular display of FRU details.

If you are going to stick with the Entity-Sensor MIB note those extreme values, like -41, -40 or -29 (different models have had diff statistic ranges, you might find a -30 something on particular models) to note where the SFP was pulled, or possibly turned off

May be we can remove the sfp port if it is not being used ?

The thing to bear in mind here is that the Hardware Sensors are not tied to monitored interfaces and therefore, evaluating the state of the interface is not going to happen automatically.

I would suggest you look to disable or change the thresholds for Warning/Critical in the Manage Hardware Sensors area. I am also as an aside, not a fan of the default hardware alert, as do you really want any warning condition to generate an alert? May be tuning the alert as well would be useful..

https://documentation.solarwinds.com/en/success_center/orionplatform/content/core-enabling-and-disabling-or-adjusting-hardware-health-monitors-for-individual-nodes-sw3596.htm#:~:text=be%20collected%20now.-,Disable%20hardware%20sensors,Management%20%3E%20Manage%20Hardware%20Sensors).

Find the sensor causing this breach in the list and adjust according to best needs to prevent this from alerting.