We have been troubleshooting an ongoing issue for the last several weeks where one of our Cisco 2901 routers intermittently becomes unreachable. When this occurs, the only resolution has been to reboot the router. At the same time, we noticed consistent high bandwidth utilization on the LAN & WAN interface of the router. Several times a day, the router would drop of the map and we would have to reboot it to restore connectivity.The router is running IOS version 15.1(4)M3. We have opened a case with TAC, but so far they haven't been much help.
After looking deeper, we noticed that the high utilization was exactly every 30 minutes. We then noticed that this 30 minute interval aligned with the default topology polling interval. At first I was very skeptical to think that was related in any way, but after changing the polling interval to 24 hours, the bandwidth spikes only occurred every 24 hours exactly.
I realize that the Topology Polling is supposedly just doing a few SNMP Get Requests, which in reality shouldn't cause any problem at all and certainly shouldn't peg a 100Mb network interface. However, I cannot argue with the fact that this is somehow directly related to the topology interval. Has anyone else experienced anything like this? I do currently have a case opened with Support.
Original Issue with Default 30-min Topology Interval:
After Disabling Toplogy Polling:
After Changing to 24-Hour Polling Interval: (WAN Interface)
