On 6/17/11 at 1:20am I had 220 of our 254 2950's (WS-C2950G-48-EI, Version 12.1(22)EA4) management interfaces disconnect. We have typical Cisco network with IP core, layer 2 HSRP distributions feeding to the edge switches. There were no network changes, no errors, no syslogs, no hsrp interface flapping, essentially nothing I could find that indicated what happened. The user vlans were still passing traffic, only the management traffic was not connecting. The management interface was up, was pingable from the switch but switch could not ping its default gateway nor could the router ping the switch's management interface. Spanning tree for the management vlan was forwarding, there were no errdisabled port.I shut down mangement interface then brought it back up.. nothing. Power cycled switch, management came up for 10 seconds or so then became unreachable again. The management interface on the switch never went down, just wouldnt pass traffic, user traffic/vlans were fine and never stopped working. I shut down the management interface, put the ip address in my laptop, configured a port on the management vlan and it worked flawlessly. Ping never failed, traffic passed for the 20 minutes of testing. Removed ip address frrom laptop, enabled management interface, nothing again. Then after being down for 6 hours all 220 switches managment interfaces came up, nothing was changed, nothing was rebooted. Another item was the closet I was in had two 2950's, same model, same software, connected to same aggregation switch, same managment vlan... one worked one didnt. Sorry for the long post but does anyone have any ideas or has anyone ran into this? Everything is still working now and I still havent found anything that leads me to anything.
