For a while now, since newer versions of v9 or v10, we have had Cisco devices randomly go from "a normal polling state" to UNKNOWN.
The UNKNOWN state clears up itself within about 30 min. We are using SNMP V2, we run "snmp-server ifindex persist" on every switch, and we have no firewalls between NPM and the switches. I even gave a higher then normal priority in QOS to SNMP to make sure it wasn't being dropped.
This rarely ever occurs on routers, instead it just affects switches (3750's, 2950's, 6500's, 4500's).
Can you give us any clue as to why this happens? I see numerous posts over the past year, but no solutions.
