IP Classless and RIP v1

[mary_odriscoll]

If I have a network with a Class C subnet on Side A communicating with a Class C subnet on Side B over a point-to-point WAN numbered with a Class B subnet address and I am running RIP V1 with IP classless configured, what happens with Side A wants to talk to Side B?

I am assuming that because RIP v1 is configured, the discontiguous subnet issue is the reason why the 2 sides can't communicate.

However, what happens with the ip classless command, is this taken into account when deciding on the routing of a packet?

[ruwhite · Accepted solution]

Do you mean two real class c networks, or just two networks with /24 masks within some larger network? RIP is very class dependant.

If you have this:

192.168.1.0/24----128.1.1.0/24--192.168.2.0/24

this really should work. If you have this:

10.1.1.0/24--128.1.0.0/16--10.1.2.0/24

it won't, with RIPv1. In the first case, you have two different major nets seperated by a portion of a third major net. In the second, you have two pieces of a single major net seperated by a portion of a second major net. In the second case, the routers on either end of 10.1.x.0/24 are going to build a summary, and advertise just the 10.0.0.0/8 to each other.

Well, it actually will work in some situations, with ip classless on, and a default route, but it's playing with fire, and I wouldn't run a network that way.

ip classless has nothing to do with the way the routing protocols propogate updates and such, it only deals with the way packets are handled when being forwarded--specifically, whether or not a supernet outside classful bounds is considered a match on the destination address for forwarding purposes.

Russ.W