I have a simple lab setup while I’m learning VyOS, three VM’s, all I’ve done is set IP addresses so there is three subnets. Each router can ping the other two directly connected routers.
If I want to configure RIP so all three routers can ping all three subnets, I would have assumed that all I should have to do is “set protocols rip redistribute connected” on each box, and they should then do exactly that, and then each router should be able to reach all subnets, I shouldn’t specifically have to advertise any networks at all.
This isn’t what happens at all, and connected networks don’t appear in any routing tables.
I have also noted that when I am advertising all networks, it does work as expected, but if I disable an interface, the interface on the other end of the cable still reports u/u, traffic flow fails and routing never reestablishes itself, I have to disconnect the cable in the VM settings for the network to converge. This doesn’t seem right at all.
Thoughts? I know RIP isn’t used much any more, I’m just using it in the lab to get the hand of the OS.
Thanks for the reply c-op, the two issues I have are:
I don’t actually know what the redistribute connected command does. Thinking about it now, I assume it is used in conjunction with “set protocols rip interface eth#” so that the connected network is advertised but the interface doesn’t actually participate in RIP?
Disabling an interface didn’t result in the other side reporting down, or in traffic being rerouted, and yes, I waited a sufficient amount of time for all the relevant timers in RIP to expire. I believe the other side remaining up is to do with the virtual environment, as it was still connected to an “internal only” (virtual box) port group.
Anyway, I’ve expanded the lab out to 6 routers and will stop wasting time on the workings of RIP, onward to OSPF, which is why I started with VyOS in the first place.