A little confusion, and I must be missing something, but here goes.
I have a customer who's main campus is connected to a remote building by T1. At each end I have a 2611 to land the T on, and connect to my network. It currently works fine with a 3COM 1100 (blech) on the remote end.
I have configured and installed a Cat3550, and when I hook it up to replace the 3COM, it comes up for a minute or 2, then the whole link goes down. I tried config with a different address and daisy-chained it off the 3COM, and it too crashed the link. What can I possibly be missing, any thoughts on what I shoudl look at.
TIA for any help
You may have looked at this already, but have you verified speed / duplex settings on the interface? I would recommend hard setting both sides and not using auto.
Indeed I did. What I find peculiar is that thing seem to work for 3 or 4 minutes and then go down. I've tried a crossover cable (didn;t work, btw), rebooting devices, the whole nine yards. It seems to me that for one reason or another, any 3550 I put in, drops the 2611 in minutes. Very odd.
Would you be able to post the output of 'show interfaces' from both the switch and the 2611 (at the time when the link is down. Also the output of 'show controllers' may be useful.
Check the remote end for redundant links.
The 3COM switches (frequently) have the setting to have "one live, one standby" line.
If you replaced that system with a switch (that didn't have spanning tree enabled), then you've got a loop ... a loop will have the symptoms you describe ... the traffic builds (storms) until the interface ODs and shuts down.
You can also just make sure the spanning tree is enabled on the 3550. If you did a "Spanning-Tree portfast" on the switch, undo it ... re-enable spanning tree and re-connect everything.
This is my show run. I will try to get the switch in tomorrow to do interfaces and controllers. This link connnects the local hospital to their transcriptionists, so I have to pick my spots cautiously. :)
Thanks again, everyone, for help with what I fancy is a small oversite by me!
Sorry for the long delay (vacation, hurray). I got back on this this a.m. Put my 3550 (102.2) back in, with port 24 configured as a trunk, with portfast. Pulled my 2611 router connection (102.1) from my old 3Com, and put it in port 24. Same result. Puke-age. Argh. I know I'm missing something simple with spanning-tree, but I'm not sure what.
Enclosed is the current show run on the 3550, and the show int on port 24 while the link was down. Actually, that's reversed, the show int appears 1st, then the show run.
Thanks again for any help.
since this post is rather long, I hope I haven't missed anything, but why is the link between the switch and the router configured as a trunk ? Do you do inter-Vlan routing on the 2611 ? Otherwise, I would suggest to configure all ports (including the port connecting to the router) as static access ports:
switchport access vlan 1
switchport access vlan 1
mls qos trust dscp
auto qos voip trust
macro description cisco-router
spanning-tree bpduguard enable
Or, you might want to remove the ' spanning-tree bpduguard enable' off the trunk port, because this will shut your port down when a BPDU is received...
Gave that a shot, nethelper, same thing, the switch port I plugged the router into errored out.
I ran this through Cisco TAC, and got a reponse to strip the 2611 fa0/0 of its programming and reprogram it with a sub-interface of 10.10.10.1, and mode dot1q encapsulation. Haven't tried it yet, but it seems quite odd to me.
It seems this should be straightforward, but man, oh, man, am I having trouble with it.
Thanks to all for the help and advice, I'll post how I turn out.