LAN DOWN - Need root cause

kamalakumar kubendran · ‎01-06-2011

Hi All,

I have faced a lan issue where i dont get any clue for the root cause. In my lan setup they are about 300 switch which consist of 6500 core switch's , 3750 \ 3560 \ 3550 distribution , 2950\2960 acess switch's. Its runnin in PVST+ as of now and it exceeded the limit of 128 VLAN as of now its runnign with 132 VLAN. Suddenly s'day all the trunk links had gone in to err-disable mode ( due to keepalive loop back ). We are in the process of migrating the lan to MST and upgradation of IOS in all the switch. We just wann to know what might have caused this issue suddenly as it was running fine with 132 VLAN for more than 3 months. We have also checked for physical loop but no such loops avail. Even in logs v dont find any mac-flap or anythin else . except a log that says loopback detected , disabling the trunk ports .

Thanks ,

Kamal

Yogesh Ramdoss · ‎01-07-2011

Kamal,

As I do not have specific details, let me give you some generic info.

An interface on a Catalyst switch is errordisabled after detecting a loopback.

Mar  7 03:20:40: %ETHCNTR-3-LOOP_BACK_DETECTED: Loop-back detected on .GigabitEthernet0/2. The port is forced to linkdown.
Mar  7 03:20:42: %LINK-5-CHANGED: Interface GigabitEthernet0/2, changed state to administratively down
Mar  7 03:20:43: %LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet0/2, changed state to down

When the peer device is going through any STP convergence, the keepalive packets sent bythe local device may get looped

back, and result in this issue.

As a result, starting in 12.2SE releases, "keepalives" are disabled by default in the fiber uplink ports. Refer CSCdz72393.

If you have Catalysts switches running 12.1 or pre-12.2SE releases, with fiber uplinks, then you may want to consider

disabling "keepalives" for these interfaces.

Hope this helps.

- Yogesh

kamalakumar kubendran · ‎12-10-2012

Hi,

Thanks for your input... Yes i know that when will keepalive packet gets looped.. but my question is how can i check which switch started this STP convergence at first .. to add i checked the TCN but no luck as it was all pointing to 10 to 15 switch's each time .. as this LAN was not properly configured (say portfast , or any other feature's).... As of the i have done with the MST migration and fixed everything ... but still i dont get the ans for my initial question of which switch started that convergence...

Anyhow thanks for all your valuable input...

Regards,

Kamal