Even though the answer is pretty much straight forward (no, you can't in most cases, but there are some exceptions!), I'm going to take a different approach and give you a bit of theory as to why you can't use certain addresses to help you truly understand the reason and not just memorize a range of numbers.
As you would already know, every network uses some type of addressing scheme. This is required in order for the computers to communicate with each other. The ranges used can vary and depend on the administrator's preference and the network's requirements. As such, most select the good old 192.168.0.0 address range, while some prefer shorter addresses like 10.0.0.0. These addresses are what we call 'logical' addresses, meaning that they are based purely on software (the operating system).
One of the very basic rules in IP networks is that every... logical network, no matter how small or big it is, must have a network identifier address and a network broadcast address. To make things simple, the engineers that designed all the above though it would be a great idea to use the first and last addresses of each logical network to do their job!
So, as you've already correctly guessed, networks ending in .0 or .255 are not usable… in most cases! The reason for this is because we tend to allocate full Class C type addresses to our networks, making it obviously very simple for us to administer.
A standard Class C network consists of 256 addresses (0 to 255 inclusive), of which one is the network address (.0) and the other is the network broadcast address (.255).
To help make this clear, if we selected network address 192.168.0.0/255.255.255.0 for our office, then we can say that:
The network identifier would be 192.168.0.0
The broadcast address would be 192.168.0.255
Valid (or usuable) ip addresses would be 192.168.0.1 to 192.168.0.254.
In the above example, addresses .0 & .255 are not usable, but in the one to follow, they are:
Consider a huge office with more than 15,000 workstations. You decide that you want all these hosts in one logical network, so a Class A network address is not able to cover your needs. A Class B gives you a total of 65,535 IP addresses – a really big network which you would surely want to break down, but for the sake of this example, we won't.
If we consider 172.16.0.0 / 255.255.0.0 as the preferred choice, then the network characteristics are as follows:
Network identifier: 172.16.0.0
Network Broadcast: 172.16.255.255
Valid (or usable) IP addresses: 172.16.0.1 to 172.16.255.254
If you take a closer look at the valid addresses, you'll notice that IP addresses 172.16.255.0 and 172.16.0.255 can be used!
The above example is the reason I didn't want you to memorize that all addresses ending in .0 or .255 cannot be assigned!
The trick to understanding all the above is of course the network theory. If you would like to read more about subnetting, IP classes e.t.c. then I'd highly advise you to visit www.Firewall.cx and check out one of the worlds finest tutorials to subnetting and IP addressing.
This was first published in January 2005