[c-nsp] Redundancy Protocol on same 7600

[Phil Mayers]

On 15/12/10 16:00, Kevin Warwashana wrote:
> I have a question regarding redundancy protocols on the 7600.  Here is my
> scenario:
>
> 1 - 7609-S
> 2 - Sup720-3BXL
> 2 - 3560's switches
>
> What I would like to do is configure a GLBP group and have both SUP's be a
> member.

There may be a misconception underlying this desire.

"Both" SUPs are not active/available for use. The 2nd SUP is in either 
cold or hot standby mode, depending on your redundancy config.

Conceptually, the device has a single active forwarding engine - and you 
can't have overlapping IPs, so basically this won't work.

> One SUP to one 3560 and the other to the second 3560.  The 3560 are
> tied together via a x-over.  So if a SUP fails the other will be able to
> route for the networks connected to it.

Again, this isn't really how the dual-sup 6500/7600 stuff works.

Just use a VLAN interface; if the SUP fails over, the 2nd SUP will take 
it on:

inteface VlanX
   ip address 1.1.1.1 255.255.255.0
int Gi1/1
   description 3560-1
   switchport
   switchport mode access
   switchport access vlan X
int Gi1/2
   description 3560-2
   switchport
   switchport mode access
   switchport access vlan X

You will need spanning tree to close the layer2 loop.

There are other ways to do this - the "backup interface" stuff just 
today discussed on the list for example.