Resilience/Redundancy on Core

[Mikecl]

I have an aging 4006 with L3 card, a 4506, and one 4503 both with different supervisor engine cards. They are connected by one link from the 4006, the floor switches then connect to either the 4006 or 4506. I would like to provide some resilince in the past I have had the luxury of having two 6500's. What is the best way (if there is one) of making the network less reliant on the 4006?

 

[dirtyhat]

Have you considered etherchanneling the Gig links between the switches? You could also add additional trunk links from the Access layer to the 4506 devices.

 

[baddos]

Also... Run HSRP between the 4506 and the old 4006. That way layer3 will also be redundant.

 

[Mikecl]

I thought the etherchannel was to help or aggregate links between devices, how does it help with resilience? Running HSRP may be an option but typically the different segments only appear on one or the other core switches. The floor switches I can run back to two core switches so that is not so much of a problem.

 

[baddos]

etherchannel protects from one link or one port going down and loosing network access, but it doesn't protect against one switch going down.