Subnetting Class C confusion

[Ballistics]

I have heard that you have to borrow a minimum of 2 bits to subnet a class C network. This would make my subnet mask: 255.255.255.192.

But I have been given a question on a quiz asking: If I have an IP address of 200.200.50.0 and a CIDR notation of /25, write down the dotted decimal format of the subnet mask.

This would be 255.255.255.128. But if what I understand is correct, how can you have 255.255.255.128 if you can't borrow just 1 bit to subnet a class C network?

Please clear this up for me. Thanks.

[sourabhagarwal]

In Class C network, you have 8 bits in host portion and 24 bits in network portion.

Subnetting is to create subnets of an exisitng network by borrowing bits from host portion. So you can create subnets by borrowing 1-6 bits from host portion because you have to leave atleast 2 bits in host portion to have atleast 2^2-2=2 hosts.

so IP address 200.200.50.0/25 will have 255.255.255.128 subnet mask.

check this URL to get more information.

http://www.cisco.com/warp/public/701/3.html

hope this answer your query .. rate if it does ...

[Ballistics]

Thank you, this answers my question.

So wherever I read that you have to borrow a minimum of 2 bits from the host portion must be wrong then, as you say we can borrow 1-6 bits.

[forkwieldingmonkey]

Previously this was the case, because the first and last network was reserved (meaning both the networks in a /25 are not usable), but cisco now allows the command: ip subnet-zero

which lets you use these two networks.

[lowes.jgibson]

That is only if you're not using IP Subnet Zero. This was typical of older routing protocols such as Rip V1. With any classless routing protocol using ip subnet zero, you can absolutely have a /25 subnet. If you used a /25 without subnet zero, that means you could not use the first or the last subnet and since there are only two subnets within a /25, there would be none left to use. That's where the borrowing at least 2 bits came from. Hope this helps.

[thomson_rc]

If you remember before ip subnet-zero, you lost the upper and lower subnets, so Cisco used to say that you had to borrow at least 2 to cover the 2 subnets lost and left you with 2 usable. This is an old rule. You can use /25.