Hig wrote:
All I'm a bit lost with is the "Mask Bits" and where you start with the IP addresses and that, such as the choice of such addresses:
172.16.0.0/27
172.16.32.0/27
172.16.64.0/27
172.16.96.0/27
oh hang on 32 is the number of subnets, so you go up in 32's ...
No, the 32 is the block size i.e. if we're going to split the third octet into 8 sections then the block size is going to be 32 (there are 8 x 32's in 256). You calculate the block size with 256-mask, i.e. 256-224 = 32.
The address range for each subnet is just the first address going up to the one before the next subnet.. e.g. for subnet 172.16.32.0/27 above we've got...
172.16.32.0
172.16.32.1
172.16.32.2
.
...missing lots out...
.
172.16.32.255
172.16.33.0
172.16.33.1
.
...missing some more out...
.
172.16.63.254
172.16.63.255
Hence, 172.16.32.0 - 172.16.63.255.
However, in practice we cannot use the first or last addresses so it becomes 172.16.32.1 - 172.16.63.254.
The /27 notation is called CIDR-notation and is just an alternative way of representing the mask.
So, for 255.255.224.0, this is written in binary as:
11111111.1111111.11100000.00000000
The first 27 bits are 1's hence we have /27.
Mathew