We are experiencing the same thing on 3750X and recently opened Case 524804. With the help of SolarWinds Support examining the snmpwalk results, it appears this is an IOS problem We are running IOS 12.2(55)SE3 on the 3750X. We haven't opened a case with Cisco yet.
Here is Support's response. Hope this is helpful.
Before proceeding with the result, let me provide you the OIDs that NPM is using to poll for the Power Supply Status/Information.
1.3.6.1.4.1.9.9.13.1.5.1.2 - ciscoEnvMonSupplyStatusDescr
1.3.6.1.4.1.9.9.13.1.5.1.3 - ciscoEnvMonSupplyState
Values:
1 : normal
2 : warning
3 : critical
4 : shutdown
5 : notPresent
6 : notFunctioning
I've checked the SNMPWalk result and found this information:
.1.3.6.1.4.1.9.9.13.1.5.1.2.1058 = STRING: "Sw1, PS1 Faulty, RPS NotExist"
.1.3.6.1.4.1.9.9.13.1.5.1.2.1059 = STRING: "Sw1, PS2 Normal, RPS NotExist"
.1.3.6.1.4.1.9.9.13.1.5.1.2.2058 = STRING: "Sw2, PS1 Normal, RPS NotExist"
.1.3.6.1.4.1.9.9.13.1.5.1.3.1058 = INTEGER: 6
.1.3.6.1.4.1.9.9.13.1.5.1.3.1059 = INTEGER: 1
.1.3.6.1.4.1.9.9.13.1.5.1.3.2058 = INTEGER: 1
Orion is only requesting SNMP request to your device and this SNMP information cannot be altered by NPM, so basically the problem appears to be on the Cisco side.
In other words it isn't Orion NPM making the determination when to warn but rather it's displaying the information it was given from the Device. So if for example the device says "The power supply is in critical state", Orion is just simply going to grab that information via SNMP and display it on web console.
I would suggest maybe checking to see if Cisco has an update to remove this information from those devices since you mentioned that the Power Supply status is in normal state. As this would also fix the problem.
Also, You do have the option to turn off health monitoring in Orion for the node however though when health monitoring is turned off you will not get any health information on that node.