[nycbug-talk] cisco switch redundancy

[Okan Demirmen]

On Fri 2005.11.11 at 14:34 -0500, N.J. Thomas wrote:
> I have a Cisco 6506 switch that I'd like to provide some redundancy for.
> 
> What I was thinking of doing was setting up a BSD/Linux box running
> VRRP, and if the 6506 died, fail over to the PC. Which will (hopefully)
> give us enough time to replace the 6506.
> 
> Has anyone done anything like this before? Is this even doable? Can BSD
> speak HSRP (Cisco's proprietary version of VRRP)?

not really.

1. cisco speaks vrrp and hsrp, openbsd speaks carp
2. are you really using the 6506 as a router *only* - that's an
expensive router, for it's really a layer 3 switch...
3. if you do any layer 2 on that and your 6506 goes, no amount of
redundancy will work if you have your devices all pluggged into that one
switch. have i seen just the rsm go in a 6500 - yes..but more often
than not, it doesn't fail gracefully like cisco'd like you to think -
usually the whole thing goes bonkers...

there are many designs on how to do redundany layer 3 switching/routing
at this scale, which are somewhat out of scope here...but if anyone
wants, we can get into it.

> Being cheap may or may not be the best option here. I know I could get
> another 6506 and then use HSRP to get redundancy, but if there is a
> cheaper solution, I'd like to hear about it first.

if you are putting out for 6506, i'm curious as to why you are worried
about $$$. redundancy at this scale can get expensive.