Subnetting for the 70-290, what do i do with the two?

Monjiro

I am a little confused about calculating the number of networks for a given subnet. Calculating the hosts is not a problem but it seems that there is two options for calculating the number of networks for a given subnet mask. The number of bits used by the mask to the power of two or the number of bits used by the mask to the power of two minus two. Both seem to be valid answers in the real world depending on best practices or the equipment you are using, what I want to know is which method does Microsoft want me to use for the exam?

Thanks in Advance

dynamik

Older equipment did not support all 0s or all 1s, so you would not be able to use those two options. That's why you have to subtract two. That is no longer an issue, so you usually do not have to do that. The question will mention (either directly or indirectly) which method to use.

Monjiro

Thanks Dynamik,

One more thing, does it make any difference the questions are talking about classless or classful ip address?

dynamik

I believe that's simply referring to whether you are using a standard A, B, or C class or whether you're working with a custom subnet.

http://en.wikipedia.org/wiki/IP_address#IPv4_address_networks

Claymoore

Even back in the NT4 TCP/IP days Microsoft supported using 'subnet-zero' by default. A true RFC-compliant subnet would lose the first and last subnet (the subnets containing .0 and .255), but when has Microsoft ever followed all the rules? Cisco follows the RFC rules by default but can be made to use the first and last subnet with the simple 'ip subnet-zero' command. For any question I have ever seen on a Microsoft exam, you can assume they are using all the available subnets.

Monjiro

Thanks very much for the help, there sure is alot of material to cover for this exam!