Applying an IP Scheme

Unanswered Question
Apr 18th, 2008

Hi guys, just a quick one. I've just purchased some kit 2 x 2501 Routers and 1 x 2924 Switch. I've been studying for me CCENT for 3 months now and I'm very close to the exam (well I think I am) and I feel comfortable with each topics... but, this threw me a bit... I was home last night designing my lab and I wasn't sure what IP scheme to play with.

I decided to go for a Class C address and a subnet in prefix length /26 which gives me 4 subets and enough host address for my setup. The way I've connected my devices gives me 3 physical segments (3 broadcast domains). 1 broadcast domain between the RouterA's Ethernet port and the Switch, another between the back to back serial link between both routers, and another between RoutersB Ethernet port and a hub. I have 2 PCs, one on the other end of the Switch and the other connected to the hub off RouterB.

So in my case I'm using, and… I've using classless routing protocols by the way.

My question is, in a real life environment, I suspect the way I've subnetted the class C address would not be ideal? If anyone can just comment on my IP setup, suggest ideal configurations that would be great.

I've done stacks of subnetting but when it comes to applying an IP scheme to a network, this is a different storey. Any advice would be great!

Thanks for reading,


  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Jon Marshall Fri, 04/18/2008 - 01:37


It really depends on what the requirements are. In my experience it is always a good idea to have at least 25% unused addressing in host/server subnets if you work in a dynamic environment because there is nothing worse than having to readdress an entire subnet.

If you are using private addressing then leaving spare capacity is not such a big deal.

For the P2P link between the 2 routers i would recommend using a /30. So with your network i would have

Always start at the end of the Class C for P2P links as it makes it more flexible or allocating VLSM addressing.

The other major thing to bear in mind when addressing is that you can summarise effectively. Using your example above again. Lets say the the P2P link connects 2 different sites. I would look to use a different Class C per site rather than split the same Class C between sites. This then gives you room for more subnets in each site out of the same Class C and you can then just advertise the whole Class C no across the link no matter how you have subnetted the Class C internally.

Edit - forgot to say, good luck with your exam when you decide to take it.


grichardson661 Fri, 04/18/2008 - 01:53

Jon, thanks for this! This is the sort of information I was looking for!!

I will change the network address for each site, this makes complete sense, specially if setting up a medium size business in 2 locations (like I am kind of…) :-)

Question regarding your /30. I thought once you have subnetted an address, I thought the subnet has to be same through each network ie 0, 64, 128, 192. Is it a case where I can just grab the 192 subnet and use /252 instead? I didn't realise this is workable?

Taking the exam around 4 weeks, should well ready then (i hope)...



Jon Marshall Fri, 04/18/2008 - 01:58


Using a different subnet mask with the same major network is called Variable Length Subnet Mask or VLSM. I don't know if this has been covered in your exam yet as i am not familiar with the exam subjects.

Classless routing protocols support VLSM so it is perfectly okay to do the following

breaking down into = -> 127 = -> 192 = -> 224


or any combination of masks within the Class C.

Classful routing protocols do not understand VLSM.


grichardson661 Fri, 04/18/2008 - 02:06

Ah, well there we go, VLSM is not covered in the CCENT but is covered in the ICND2 and CCNA, hence why I questioned your /30 before.

I understand the classless and classful routing protocols but not VLSM yet! Anyway, thanks again!



This Discussion