One other note, rfc 1878's recommendation allows for the usage of the first
and last subnet. Routing protocols that carry subnet information in its
updates allow for this. Ones that don't (igrp, static, ripv1, etc.). The
practice I follow is to use them last, if needed.
Charles
At 01:55 AM 10/1/96 +2600, Harry Feltsadas wrote:
>John McColley @ J F Engineering wrote:
>>
>> Let's see if I understand subnetting correctly.
>> If I want to split a class C network to setup a firewall I would
>> take the existing network, say a.b.c.0 with a netmask of
>> 255.255.255.0 and instead I would end up with 2 usable networks
>> if I use a netmask of 255.255.255.192. I would end up with
>> network a.b.c.64 with a netmask of 255.255.255.192 and network
>> a.b.c.128 with a netmask of 255.255.255.192. Therefore, I would
>> have available addresses of a.b.c.65 (netmask 255.255.255.192)
>> through a.b.c.126, broadcast address would be a.b.c.127 and
>> a.b.c.129 (netmask 255.255.255.192) through a.b.c.191, broadcast
>> address would be a.b.c.192.
>> I can't use a.b.c.0 through a.b.c.63 and a.b.c.193 through
>> a.b.c.255.
>> Does this sound right?
>>
>More or less. Subnetting uses a subnet mask. Any bit that is on in
>the mask is a network bit. If the bit in the mask is off, the
>bit belongs to a host part of the address. In your case above,
>using a mask of 192 means that the first two bits are the subnet,
>the rest are hosts. The first two bits give 4 networks (not two).
>So in this case, you can use all four!
>If you want two subnets, use a different subnet mask, and remember,
>the net bit can be anywhere, so you can divide it into two parts,
>as long as it's a power of two. This way, you don't lose that
>many numbers!
>
>--
> ) _) Harry Feltsadas __
> / / mailto:harry @
fdc .
nl __ __(_ )
> _ ) _ _/ http://www.fdc.nl/~harry ' | '/ /
> __/ __/ __/ < A Red Goose and a Red Duck > ___/ ___/
>
>
Follow-Ups:
|
|
