Classful Routing

I understand the difference btw classful and classless (well - at least I think I do)

But I came upon an example reading today that has me confused.

Lets say that you're using RIPv1 and have a configuration like something below

RouterA
int Serial 0/0/0
ip address 172.16.4.33 255.255.255.224
encapsulation ppp
no shut

router rip
network 172.16.0.0

RouterB
int Serial 0/0/0
ip address 172.16.4.34 255.255.255.224
encapsulation ppp
no shut

router rip
network 172.16.0.0

I might be missing a thing here or there with that config - just did it from memory - but the main question I have is that will RIPv1 advertise subnet 172.16.0.0 with a /16 or a /27

I always thought that classfull protocols assumed the normal class network mask /8 /16 /24. But the example I saw looks like RIP got the mask from the interface and assumes that for all networks?? Not sure there. Any help would be great.

Find more posts tagged with

Save $250 on 2025 certification boot camps from Infosec!

Book now with code EOY2025

Button

Comments

Webmaster

I understand the difference btw classful and classless (well - at least I think I do)

The following shows you are still missing an essential part:

but the main question I have is that will RIPv1 advertise subnet 172.16.0.0 with a /16 or a /27

RIPv1/classful protocols, do not advertise any subnet info. There isn't a field for a mask in the RIPv1 update packet. So it has to determine the mask using a different method.

I always thought that classfull protocols assumed the normal class network mask /8 /16 /24

Only when it cross a major network boundary. So if you would connect a router with 10.0.0.0 network behind it, routers on that side would receive 172.16.0.0 with no mask info, and since they are not in the same network, they don't know which mask was used on the interface that originally send the update, hence would assume /16.

If they would always use /8 /16 /24, you wouldn't be able to subnet a network that uses RIPv1, which is not the case, the subnets just have to be of the same length (hence no VLSM), and contiguous . See the following docs for an example network showing why VLSM and discontiguous would mess up a RIPv1 or IGRP network:

www.cisco.com/en/US/tech/tk365/technologies_tech_note09186a0080093fd6.shtml

www.cisco.com/en/US/tech/tk364/technologies_tech_note09186a0080093f1e.shtml

bmauro

I think I worded my question wrong.

I understand that RIPv1 does NOT transmit subnet information at all with its routing table updates. So forget what I asked about that above.

I'm having a hard time expressing this question - So I'll continue reading and hopefully figure this out.

Just bought TCP/IP Routing Vol1 by Jeff Doyle - so if I can't find my answer in there - then I'm in trouble

bmauro

Johan - i was writting my reply before I saw you posted - but I'm reading the Cisco doc right now - and I think you "got" what I was trying to ask.

Thanks

Webmaster

Yeah, you can't really miss "does not include subnet info" when you read about RIPv1, IGRP / classful routing protocols, so I figured I'd just focus on your second question.

bmauro

Alright - after reading Jeff Doyle's Routing TCP/IP for 10 minutes I can say that if ANYONE has any questions or wants a deeper understanding of interior routing protocols - get this book. Its for CCIE canidates - but WOW, the infotmation in this book is amazing.

Johan - you got my question. How can you subnet if RIPv1 doesn't include masks with updates. You would think that routing tables would only have entires using the classful addresses (at least thats were I was wondering).

This part from Doyle's book completely cleared it up.

"If the router whose forwarding database is depicted in 5-2 receives a packet with a destination address of 172.25.131.23 /24, there is no positive way to determine where the subnet bits end and the host bits begin, or even if the address is subnetted at all.

The router's only recourse is to assume that the mask configured on one of its interfaces attached to 172.25.0.0 is used consistently throughout the network."

Of coarse the book goes on showing examples of discontigous networks (why they don't work) just like the Cisco.com examples posted earlier (thanks again Johan).

Webmaster

A bit too detailed for the CCNA exam, but it certainly doesn't hurt to read a CCIE level book.

And you're welcome, I'm sure your topic will be useful for others as well.

lon21

bmauro wrote: »

Alright - after reading Jeff Doyle's Routing TCP/IP for 10 minutes I can say that if ANYONE has any questions or wants a deeper understanding of interior routing protocols - get this book. Its for CCIE canidates - but WOW, the infotmation in this book is amazing.

Johan - you got my question. How can you subnet if RIPv1 doesn't include masks with updates. You would think that routing tables would only have entires using the classful addresses (at least thats were I was wondering).

This part from Doyle's book completely cleared it up.

"If the router whose forwarding database is depicted in 5-2 receives a packet with a destination address of 172.25.131.23 /24, there is no positive way to determine where the subnet bits end and the host bits begin, or even if the address is subnetted at all.

The router's only recourse is to assume that the mask configured on one of its interfaces attached to 172.25.0.0 is used consistently throughout the network."

Of coarse the book goes on showing examples of discontigous networks (why they don't work) just like the Cisco.com examples posted earlier (thanks again Johan).

Hi is this correct?

When using Classful routing, the router uses the default subnet of the first octet IP address to determine its subnet mask. As subnet info is not sent?

E.g.

10.1.1.0
Would be 10.0.0.0 as class A subnet is 255.0.0.0

192.168.2.10
Would be 192.160.2.0 as class C subnet is 255.255.255.0

Also is it correct to say that in the exam we should not think of auto summarzation and always summarise the network to there classful mask i.e. 82.3.5.23 should be 'network 80.0.0.0'

Thanks