Beginner needs help with CIDR

[ChuckCasey]

We have a class c 192.168.1.0 network that is running out of ip addresses. We have 4 branches with routers at each one for the T-1 connections. I thought that I could supernet to get 192.168.0.0 - 192.168.3.255 with a subnet of 255.255.252.0. I'm not sure what configuration changes I would have to make on the routers to keep all 4 branches talking. I assume I need to change the subnet mask. Do I have to add route statements, or is it unnecessary because they will still be in the same subnet? Please advise. Thanks so much.

 

[chieftan999]

I may be wrong with this but here goes:-

Because you are using a 192 address that is a class C.

A Class C default subnet is 255.255.255.0 therefore the network address is the first 24 bits (3 Bytes) and the last 8 bits are reserved for hosts.

Subnetting is borrowing bits from the host portion.

You cannot have a subnet of 255.255.252.0 with a class C address. I think you need to look at the following:-

Class A - Network.Node.Node.Node
Class B - Network.Network.Node.Node
Class C - Network.Network.Network.Node

Range Class A = 00000000 = 0 (Must be 1)
01111111 = 127 (127 used for loopback. 126)

Range Class B = 10000000 = 128
10111111 = 191

Range Class C = 11000000 = 192
11011111 = 223

If we start borrowing bits from the host on a class C we get the following:-

First LEGAL Subnet Class C = 11000000
This gives us a subnet mask of 255.255.255.192 which gives 2 subnets of 64 and 128.

The math is as follows:-
For subnets available for a given mask then look at all the bits switched on (1) and then power them (in this case 2 to the power of 2 - 2 = 2 (64 and 128) and the hosts are the bits switched off (6 to the power of 2 - 2 = 62).

Next subnet mask is 11100000 = 224
Then 11110000
then 11111000
then 11111100

That is the last one since there must be 2 bits available for the host.

I think you may be looking for a class B subnet mask and address scheme.

Under the terms as stated above on a class C you would have to use 255.255.255.224 to allow for 6 subnets (you require 4) but that only gives 30 available node addresses per subnet. If that is okay then go ahead and do it. If it is not enough addresses then you will have to go with a class B.

 

[bierhunter]

Yes, with CIDR, you can use 192.168.0.0 /22. As long as your equipment supports classless (most do), you should be fine.

You will end up with

192.168.0.0 - 192.168.3.255
Mask: 255.255.252.0

192.168.0.0 = network address
192.168.3.255 = broadcast

All others are usable host addresses.

That should work fine, people are doing it everywhere.

Just a helpful tip though. This is just my personal preference. I would stay away from the 192.168.x.x for company networks. Only because most home routers use those ranges. If you ever have any remote vpn users, you could end up with subnet overlaps.

I just stick to 10.x.x.x or 172.16.x.x - 172.31.x.x for business use and leave the 192.168.x.x for the home users. It's saved me some hassles with our vpn users.

Hope this helped.

BierHunter
CNE, MCSE, CCNP

 

[ChuckCasey]

Many thanks for the help. I was able to change the subnet mask and have everything connect except for one thing. We have a vms system that I cannot connect to from any of the new addresses (192.168.0.x, 2.x, 3.x, 1.x still works fine). The vms system has a 192.168.1.x address. I'm not very familiar with this system, so if anyone has any ideas, please feel free to offer them.
______________________________________________________
Just a helpful tip though. This is just my personal preference. I would stay away from the 192.168.x.x for company networks. Only because most home routers use those ranges. If you ever have any remote vpn users, you could end up with subnet overlaps.
________________________________________________________

We'll probably go this route in the future, bierhunter. My main concern at this time is keep as far away from the VMS system and DEC servers as possible until I'm better trained.

 

[bierhunter]

If the VMS system is old enough, it may not recognize the /22 subnet mask.

I've seen some legacy equipment that will only use classful networking. That may be the problem in this case. The VMS may be forcing a /24 subnet mask.

In that case, you may have to have multiple subnets on your interface to compensate for the VMS. I'm speaking from a Cisco point of view, but I don't know what you're using.

For example:
interface FastEthernet 0/0
ip address 192.168.0.1 255.255.255.0
ip address 192.168.1.1 255.255.255.0 secondary
ip address 192.168.2.1 255.255.255.0 secondary
ip address 192.168.3.1 255.255.255.0 secondary

But...if you can get the VMS to recognize the /22 subnet mask, you should be fine. Legacy stuff can be a pain sometimes, but there is still a bunch out there.

Also, to help solve the legacy equipment problem (if that is the case), you could switch to 172.16.0.0/16 and just use it as a class-B. The legacy system should work with that and you'll have bunches of addresses.

Interface FastEthernet 0/0
ip address 172.16.0.1 255.255.0.0

This will of course require resubnetting everything, but it sounds like you're in the middle of that anyway.

BierHunter
CNE, MCSE, CCNP

 

[rcpis]

Our network, as well, is running out of IP addresses. We number it 10.100.10.x with a 255.255.255.0 subnet. I want to do the same as well, change the subnet, resubnet, to give me more IP addresses. If I change the subnet to 255.255.254.0 do I end up with 10.100.9.1 - 10.100.10.254 from which to pull addresses from? I find this very confusing. Thanks for the help.