Spanning-Tree config for a looped cable LAN

[maverick-lamont]

Hi,

Hopoing someone could help clarify if the below config could work,  or if it is just a no no.

I am dealing with 2 DC's operating with a stretcheld LAN configuration using plain dot1q. 

We have a 1GB LAN link between the 2 of them and an older 100MB LAN link.

Can this configuration work:

DC1-CORE-1 - 1GB LAN Link - DC2-CORE-1

DC1-CORE-1 - 1GB - DC1-CORE-2

DC1-CORE2 - 100MB LAN Link - DC2-CORE2

DC2-CORE1 - 1GB - DC2-CORE2

It truly is a looped LAN configuration but we want to the 100MB to just act as a secondary path in the event a 1GB goes down.

Let me know your thoughts.

Thanks

[paul driver]

Hello
stp port priority and cost should converge to the 1gb links.

Can you post he out put from

Sh trunks
Sh span int xxx


Res
Paul

Sent from Cisco Technical Support iPad App

[maverick-lamont]

Hi Paul,

At the moment the 100MB is disconnected becuase the loop crashed the network even with the spanning-tree setup.  I am not sure how usefull those outputs will be with the 100MB disconnected and I would prefer not to post that info. 

Basically the 100MB had the same priority as the 1GB links but thre was a higher cost for them so I would expect the spanning tree to do the job and block over the 100MB. 

Any ideas what could be the issue?

Thanks

David

[Jon Marshall]

David

I'm sure you haven't so please don't take offence but the ports don't have portfast on them do they ?

Also, have you changed any port costs or have you left them at the default ?

STP should block on one of the 100Mbps ports. Which one depends on which switch is the root bridge for the vlans.

Are these trunk links or access port links. If trunks can you confirm the native vlan is the same on both ends of the trunk ? Even if they are not it should not cause a loop as STP shoul shut one of the ports down but worth checking.

What was the sequence of events ie. the minute you brought the interfaces up did it crash the network or was it up and running for a while ?

Which version of STP are you running ?

I appreciate it is difficult with it disconnected but to the best of your knowledge is the 100Mbps link working properly in terms of bidirectional traffic ?

It's difficult to be more precise because as you say any STP outputs now wouid not be that helpful.

Jon

[maverick-lamont]

Hey I think I have found the problem.

I can see CORE-1 in both DC's are root for the VLAN's.

And the trunk between dc2-core-1 and dc2-core-2 is ISL and the rest are 802.1q

Will correct and then I think it will come right and let you know.

Thanks

David

[paul driver]

Hello
have you checked the cabling?
Do you have any layer one device (hub) attached to either of these switches which could possibly introduced a loop?
Do you have any old span sessions related this these ports?
Is stp being disabled via bpdu filters
Are these trunks and if so are they set to negotiate trunking?
Are these ports aggregated?
Could you post the config for these ports?

Res
Paul



Sent from Cisco Technical Support iPad App

[devils_advocate]

To clarify, this is your setup and each of the links are dot1q trunks?

Are you running Rapid-PVST?

Assuming you have not changed any of the costs or the port priorities then the 100mb link should be put into a blocking state.

Can you post the output from each of the interfaces that interconnect the switches?

Thanks