RIP vs RIPv2

alliasneoalliasneo Posts: 186Member

Hi, I'm a little confused over RIP vs RIPv2.

I have read over and over but can't find a definitive answer in any of the books I'm reading.

My question is, what routes does RIPv1 route too? I've included a sample of my lab - which I hope everyone can see. If RIP is classful and I'm using 168.15 then RIPv1 automatically thinks 255.255.0.0 right?

But if I've subnetted down like I have to .252 to only allow 2 numbers per network on the serial interfaces, how does this affect routing? I'm just trying to see why version 1 is limited to such a degree? I can't quite see the reasoning behind the routing?

So if I had another network off of R2 with the same network 168.15 but on a different subnet number would the routing not work for packets that get to it? Would they be forwarded out the serial interface going to R1 as well as the other network?

-I hope this is clear.

thanks
lab.jpg 43.5K

Comments

  • alliasneoalliasneo Posts: 186Member
    Sorry to be premature guys, I was just looking at this and have examined the routing tables and have come up with the following explanation:


    R3 thinks it has routes to the entire 150.150 network - because 150 is a class B
    this would be /16 (150.150.0.1 - 150.150.255.254).


    So if you try to ping 150.150.2.1 it will fail because the 150 network is subnetted down to /24


    If I migrate to RIPv2 this should resolve this by sending the subnet mask out and R3
    will then know that the 150 route is 150.150.1.0 with 254 devices.


    Would this be right?
  • sasprosaspro Posts: 114Member
    RIPv2 supports vlsm so would sort this out for you.
  • MrBrianMrBrian Posts: 520Member
    The main differences are this:

    -RIPv1: Sends updates to 255.255.255.255(broadcasts). And it does not send the subnet mask information.
    -RIPv2: Sends updates to 224.0.0.9(multicasts). And it does send the subnet mask for each prefix.

    With RIPv1 you can actually use subnets, but you're walking on eggshells. RIPv1 won't send any subnet info at all, but it can send something like 172.16.20.0, without summarizing to 172.16.0.0. It will just send it as 172.16.20.0 with no subnet info. However the RIP enabled interface will only send it like this if you're sending the updates over a link that's apart of the same classfull network, and using the same mask length. For example, if you have class B subnets 172.16.20.0/24 and 172.16.30.0/24 on one router, and the link connecting to another router is 172.16.10.0/24, they won't be auto-summarized, they'll be sent as 172.16.20.0 and 172.16.30.0, but with no subnet mask info. The other side will put them as /24's in its routing table cause it assumes they're the same. If the link connecting the routers had a different subnet mask though, like a 255.255.255.252, then the router with the subnets will suppress it's updates accross the link because the interface it wants to send the updates out of has a different mask, even though they're in the same classfull net of 172.16.0.0. Otherwise, if the links in between are of different classfull networks, then RIPv1 will just summarize the subnets to their classfull boundary and send them. Basically, you can't use RIPv1 if you want to use the same classfull network with different subnet masks like /24 here, /30 there (VLSM).

    With RIPv2 however, it will send the prefix AND the subnet for its updates. This allows you to have networks using varying length subnet masks (VLSM). /24's will get sent over a link using /30 because the other side will be able to know it's a different subnet. Umm, I hope I explained all that somewhat decently lol. Also, here's a good link for how RIPv1 sends and receives updates: Behavior of RIP and IGRP When Sending and Receiving Updates - Cisco Systems
    Currently reading: Internet Routing Architectures by Halabi
  • thedramathedrama Posts: 291Member
    alliasneo wrote: »
    Hi, I'm a little confused over RIP vs RIPv2.

    I have read over and over but can't find a definitive answer in any of the books I'm reading.

    My question is, what routes does RIPv1 route too? I've included a sample of my lab - which I hope everyone can see. If RIP is classful and I'm using 168.15 then RIPv1 automatically thinks 255.255.0.0 right?

    But if I've subnetted down like I have to .252 to only allow 2 numbers per network on the serial interfaces, how does this affect routing? I'm just trying to see why version 1 is limited to such a degree? I can't quite see the reasoning behind the routing?

    So if I had another network off of R2 with the same network 168.15 but on a different subnet number would the routing not work for packets that get to it? Would they be forwarded out the serial interface going to R1 as well as the other network?

    -I hope this is clear.

    thanks

    I consider your questions do not have an answer. i mean these two routing protocols had been developed to work
    differently only. Developers thought that RIP may operate like this and lets make it different on RIPv2 and so on.
    Monster PC specs(Packard Bell VR46) : Intel Celeron Dual-Core 1.2 GHz CPU , 4096 MB DDR3 RAM, Intel Media Graphics (R) 4 Family with IntelGMA 4500 M HD graphics. :lol:

    5 year-old laptop PC specs(Toshiba Satellite A210) : AMD Athlon 64 x2 1.9 GHz CPU, ATI Radeon X1200 128 MB Video Memory graphics card, 3072 MB 667 Mhz DDR2 RAM. (1 stick 2 gigabytes and 1 stick 1 gigabytes)


  • alliasneoalliasneo Posts: 186Member
    thanks everyone for the reply's. Hmmm it's not easy to understand. I've done a little on VLSM but I don't hit that properly until ICND2 so I'm wondering if I shouldn't worry about it to much for now? I'll keep at it though and see. I'm just finding it had to determine what the router will advertise and how to figure this out?

    One example I had was a question of which route will be chosen based on this mac address table and the following packets.

    I just don't understand at the moment how you figure out the most specific route based on this information?
  • alliasneoalliasneo Posts: 186Member
    OK I've revised my network slightly and added a 150.150.1.0 network with a /24 mask.

    Now RIPv1 send 150.150.0.0 - because it is auto summarising it for a class B?
    So Router 3 receives this update, adds it to the routing table and thinks it can reach the whole 150.150 network?
    However, when R3 tries to ping one of these networks it gets back to R1 and R1 drops it? I got a u.u.u as R3 thinks it has the route but R2 doesn't.

    Any of this correct?

    Thanks
  • MrBrianMrBrian Posts: 520Member
    alliasneo wrote: »
    OK I've revised my network slightly and added a 150.150.1.0 network with a /24 mask.

    Now RIPv1 send 150.150.0.0 - because it is auto summarising it for a class B?
    So Router 3 receives this update, adds it to the routing table and thinks it can reach the whole 150.150 network?
    However, when R3 tries to ping one of these networks it gets back to R1 and R1 drops it? I got a u.u.u as R3 thinks it has the route but R2 doesn't.

    Any of this correct?

    Thanks

    You're correct, if using RIPv1 it will send the advertisements as 150.150.0.0, summarizing to the classful net. And yes, Router 3 will think it can reach the entire 150.150.0.0/16 range now. When you ping and receive u.u.u those are responses from another router saying the address was unreachable.. so the pings are being sent to R1, but for whatever reason R1 can't reach 150.150.1.1, so it sends back ICMP destination unreachable error messages which are the U's you're getting on R3.

    On your 150.150.1.1 host did you set its default gateway? This should be the IP that you have configured on R1's LAN interface. Then the pings should come back to R3 ok. Let me know if you have any problems
    Currently reading: Internet Routing Architectures by Halabi
  • alliasneoalliasneo Posts: 186Member
    thats brilliant. Yup it all seems to be working now and I have migrated it over to RIPv2 and the routing table now looks like this:

    C 1.1.0.0 is directly connected, Loopback0
    150.150.0.0/24 is subnetted, 1 subnets
    C 150.150.1.0 is directly connected, FastEthernet2/0
    168.15.0.0/30 is subnetted, 3 subnets
    C 168.15.240.244 is directly connected, Serial1/0
    R 168.15.240.248 [120/1] via 168.15.240.253, 00:00:28, Serial0/0
    [120/1] via 168.15.240.246, 00:00:14, Serial1/0
    C 168.15.240.252 is directly connected, Serial0/0

    I can see now that RIPv2 has sent the /24 info and is only advertising routes for the correct network and I can ping this successfully from each router.

    Thanks all for your help. I'm still a little unclear about routing for the best path but I'll post something else up on that when I have looked into it a bit more.
Sign In or Register to comment.