Type I cabeling,Hermaphrodical short circuit prevention ?

fmerkle · ‎01-09-2004

Hi Cisco Team,

we are using type I (TokenRing) cabling in our

campus. Hermaphrodical connector to integrated RJ45.We migrated to ethernet/fastethernet cisco components a few years ago. Sometimes we have the problem of short circuits on the Hermaphrodical connectors, and this brings major problems to the attached switches. Sometimes it happens on switchports who are not connected to end devices but they are attached to the patch pannel with the Hermaphrodical connector. when a short circuit happens on the switch the error count of the port for the in/out frames are raising very fast and after a short period the whole switch goes down.

Is there a way to prevent those problems or are there ways to enable trap messages for this case ?

Maby a kind of a counter who will block the port when the allowed error count exceeds the limitations ?

We use in our campus cisco 3***,4***,5***,65**,85**

switches and we use CW2000.

Any suggestions are welcome :-)

Thanks for your help

Regards,

Frank

Georg Pauwen · ‎01-09-2004

Hello Frank,

you could try and enable environmental SNMP traps with the command:

snmp-server enable traps envmon

And instead of the entire switch going down, you could configure storm control for the ports on the switch, with the command:

storm-control {{broadcast | multicast | unicast} level level [lower-level] | action shutdown}

Regards.

GP

lnguye · ‎02-24-2004

The solution is to

1) turn on BPDU guard (set spantree portfast bpdu-guard enable),

2) set the access ports as host port (set port host /). Note: never set the switch-to-switch port as host port or portfast.

3) and make sure the switch would not automatically re-enable the disabled port due to BPDU errors (set errdisable-timeout disable bpdu-guard)

Just remember, someone will have to manually enable the port if it was disabled previously by the BPDU guard process, but that's better than affecting other traffic, or crash the network.

LQN