Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 

Welcome to Cisco Support Community. We would love to have your feedback.

For an introduction to the new site, click here. If you'd prefer to explore, try our test area to get started. And see here for current known issues.

New Member

Cat 6500's Running Native in a Distributed Bridge Design

I have several Cat6500's (10 as of last week) that are installed with SupII's, MFSC's and PFC2 cards running Native IOS 12.1.13E code. There is one vtp server in my vtp domain and the network typology is a hub and spoke. The network has been segregated into several VLAN's. Each VLAN is specific to the area the switch is responsible for, e.g. VLAN 10, 11 and 12 is designated for switch 1, vlan 13, 14, and 15 is for switch 2.

From a design theory question, should I place the root bridge for each of these vlans on their respective switch, having the core act as secondary or collapse it down into the central core switch that is also the vtp server?

What do you think? Thanks for your input on the matter.


Re: Cat 6500's Running Native in a Distributed Bridge Design

When you specify a different root for each part or vlan, you will achieve higher fault tolerance. I would prefer this design over having one root for all vlans with the main argument that the whole network will become unstable when something happens to the core switch.

For choosing a backup-root, it is important to know how the data is supposed to flow in case of a root-failure. In this situation you should pick a device that remains reachable for the entire affected vlan. This should not nescessarily be the core-switch.

Re: Cat 6500's Running Native in a Distributed Bridge Design

This looks like a bad design. Without a more detailed look at what you have got and what you want to achieve I would suggest looking at redesigning the network and removing 'spanning' VLANs as much as possible. If possible look at having 'pairs' of Cat6500's with common VLAN's spanning the two and using HSRP between them, then interlink the pairs of switches to other pairs using Layer-3 point-to-point (/30) networks. This way you will have the resilience but not the potential melt-down due to STP breakdowns. I would also avoid using VTP; its a handy protocol but is it really needed?