[c-nsp] HSRP vs VRRP

Tim Durack tdurack at gmail.com
Thu Oct 20 14:21:39 EDT 2005


> So I need to ask my question again: how do you *do* this on the routers?
>
> To make HSRP work, all interfaces need to be in the *same* Layer3 network
> segment - and Cisco routers are fairly unwilling to tolerate that, unless
> you do etherchanneling (or "backup interface") stuff.
>
> We're not talking "L3 capable switches and HSRP is run on the MSFC" here
> - plain L2 switches, plain L3 routers.
>
> gert

RTR-1 connects to SW-1, x.x.x.3 is the interface address, x.x.x.1 is
the standby ip.
RTR-2 connects to SW-1, x.x.x.4 is the interface address, x.x.x.1 is
the standby ip.

RTR-1 connects to SW-2, y.y.y.3 is the interface address, y.y.y.1 is
the standby ip.
RTR-2 connects to SW-2, y.y.y.4 is the interface address, y.y.y.1 is
the standby ip.

Standard L3 routers and L2 switches employed.

You can configure .2 as a second backup group if you want some
rudimentary client load-balancing. Or run GLBP instead :-)

Maybe we are talking at cross-purposes here. I was just mentioning
that Cisco don't really recommend what you initially posted because of
the split switch problem you mentioned.

It is a recognised weakness/limitation/feature or however you like to
describe it...

Tim:>



More information about the cisco-nsp mailing list