The Art IP Subnetting 1

[slycer]

Im not sure if I am looking in the right forum, but I figured I would ask. Is there anywhere on the net that will walk you through how to Subnet IP addresses?

 

[bfjacobs]

You might try

http://www.learntosubnet.com/

 

[Ntr0P]

Also,

http://www.subnetonline.com.

Haven't used it so I can't endorse it, but I have seen it mentioned a few times in response to similar questions.

 

[chieftan]

There is a very easy way to complete this operation:-

The first is how the class ranges are worked out.
Class A addresses cannot have the first bit set to "on" so if you turn all the other bits on and then off you have your class A range:-
00000001 = 1
01111111 = 127 (This is used for loopback however)

Class B - First bit MUST be on and second bit MUST be off. Turn all the other off and then on and you have the class B range:-
10000000 = 128
10111111 = 191

Class C - First 2 bits MUST be on and the third bit MUST be off. Turn all the other bits off and then on and you get the class C range of addresses:-
11000000 = 192
11011111 = 223

The other IP addresses above this are reserved for testing purposes.

Subnetting:-

The art of borrowing bits from the host portion. The part that most people have the problem with is knowing the amount of hosts and subnet that will be available with each mask. This is easily solved if you know your powers of 2.

In the following example pretend we are using a class C address and Subnet mask. Lets say the first byte is 198. The default for the class C subnet is 255.255.255.0. This leaves 8 bits for the host portion (254 addresses). We cannot borrow 1 bit from the host portion as with a class c this is illegal so we have to borrow 2 as a minimum. This means the first subnet mask we can have is 192 as follows:-

255.255.255.192 = 11111111.11111111.11111111.11000000.
Now, how do we find out how many subnets that will give us and how many hosts per subnet:-

1: Use the power of 2 for all the ON bits and then -2 at the end. This will equal 2 (2 to the power of 2 = 4 -2 = 2).
2: Use the power of 2 for all the OFF bits to find the amount of hosts. This will equal 62 (6 to the power of 2 = 64 - 2 = 62).

Do the same format with each borrowed bit. I'll show you one more.

255.255.255.224 = 11111111.11111111.11111111.11100000

Subnets = 3 to the power of 2 = 8 - 2 = 6.
Hosts = 5 to the power of 2 = 32 - 2 = 30.
So in this case we have 6 subnets and 30 hosts per subnet.

 

[svermill]

Chieftan,

That's a very nice detailed example. Is Cisco Networking Academy still teaching the -2 subnets? 'Cause the all-ones subnet was always valid in real networks (with some very minor exceptions as I understand it) and the all-zeros subnet has been enabled by default on Cisco IOS for years. I don't think the -2 part is valid anymore, but I don't know what Cisco exam writers are expecting people to do to get the questions right.

Thanks.

 

[Ntr0P]

Yes, the -2 subnets is still alive in well in the minds of Cisco. Don't know why since ip subnet-zero is enabled by default as you pointed out.

 

[bierhunter]

Mainly they teach the -2 just for the tests. Cisco fully understands that in the real world, we use the full range of subnets. But the RFC "standards" still have some -2 in them. So, for testing purposes, they want the people to understand the "standard" way of subnetting.

It's getting real ancient though. I don't know of any manufacturers that don't use subnet zero anymore.