subnet masks

Dear Tarun, i have elaborated the full 8 "bits" on this issue, no pun intended! hope it aids your comprehension.

I still don't understand which "8 bits" are you talking about because it depends on the class of IP Address you are talking about, if it is a class "A" IP Address then the host part has 24 bits so you would have to specify which "8 bits" are you talking about, also if it is a class "B" IP Address then the host part has 16 nits so again you would have to specify which "8 bits" are you talking about.

So, its not really clear what are you looking for.

And yes, if it was clear i am sure we have a lot of guys who are more than willing to help you out, but only if they could understand what the question was :lol:

And yes NO PUN INTENDED

It seems that Arani is having problems understanding only the Class C classfull ips as its the only Class left that has 8 bits in the host part... :roll: still if u can elaborate the question shurely we will come up with some clear answers

do all of them use the first 3 bits (from the left) of their respective host ID's to assign subnet masks?

Why only the first 3 bits, it can be as many bits as you want up to a maximum of /30 which would give you only 2 usuable hosts

None of the classes of IP’s are restricted to subnetting by way of using just the first 3 host bits. Subnetting in all classes requires that you use at least 2 bits but you must also leave 2 bits. That means starting from the left most host bit:

In class A that is 2 – 22 bits
In class B that is 2 – 14 bits
In class C that is 2 – 6 bits

None of the classes of IP’s are restricted to subnetting by way of using just the first 3 host bits. Subnetting in all classes requires that you use at least 2 bits but you must also leave 2 bits. That means starting from the left most host bit:

In class A that is 2 – 22 bits
In class B that is 2 – 14 bits
In class C that is 2 – 6 bits

Just a word of caution here when you say that it requires at least 2 bits for subnetting. When you say that, i guess you are talking about "cisco" specifically because otherwise subnet's can be made with just one bit being used for subnetting with a corresponding subnet mask of 255.255.128.0 for class B & similarly for other classes.

i thought for a while that there was some rule in deciding how many bets we should be using to decide on subnet masks. i was swayed by the example that was used to describe the concept on subnetting, and kept on thinking that we have some rule on deciding how many bits from the host id we will be using. and me harping constantly on "8 bits" was because i was stuck on the class C addresses for some time. but let me clear some air about it now. any address, any class, we use the host id portion no matter how long it is, for subnetting. we can take as many bits from the host id we want to make up a subnet mask. but there is some consideration that has to be made regarding how many networks we want and how many hosts per network we need to setup. that is the guiding light to how many bits we will be needing from the host id portion. this is irrespective of the class (i.e A, B or C).
I sincerely hope i have made my question and my answer clear as well. if i am wrong anywhere..please do feel free to correct me