RIP passive interface

[richmorrow624]

I have two 4503 switches connected via etherchannel and configured for HSRP.

They are both running RIP, righ now the switches are configured with six secondary interfaces all in vlan1.

Until the time when I can reconfigure the flat network, is it possible to apply the passive interface to the secondary addresses and reduce some of the RIP traffic?

[jackyoung]

AFAIK, the "passive interface" is applied to the interface-based and not address-based. Therefore, if you are using VLAN to for each subnet (each secondary IP) then it will be possible. However, you are using secondary IP in a SINGLE interface then if apply passive interface, it will passive the whole interface intead of individual subnet.

Moreover, it is not necessary to passive the secondary address. It was because IF you can do it, but it is still using the same physical interface for other ACTIVE address, so the routing protocol will still be carried in the same physical interface, no matter you passive the address or not.

Will you consider to reconfigure the network as VLAN-based ? Unless there are multi-address in host, otherwise, VLAN should be possible to implement in the network.

Hope this helps.

[Richard Burts]

Richard

I take a slightly different approach than Jack but come up with the same answer.

There is not any option to do passive interface for a secondary address. (The command is not passive-address it is passive-interface and as Jack observes it affects the entire physical interface).

Beyond that I think that the logic that thinks that it will reduce RIP traffic is flawed. If you assume that RIP will send an update for each address it would make sense to try to passive the addresses. But RIP sends ONE update per interface (sourced from the address of the primary address). The only impact of secondary addresses is the number of networks/subnets that RIP will advertise. (And do be careful that your RIP network statements do cover all of the addresses that are secondary).

HTH

Rick

[jackyoung]

Thanks. Rick, for your comment. :)