Subnetting
By Victor Napolillo, ecpi Technical College
Each and every IP address that you will ever work with will fall into one of several classes, but always one. Each of these classes has its own default subnet mask that is used if no subnetting is done:
Class IP Range Default Subnet Mask High Order Bits
Class A 0.x.x.x – 126.0.0.0 255.0.0.0 0
Class B 128.0.x.x – 191.255.0.0 255.255.0.0 10
Class C 192.0.0.x – 223.255.255.0 255.255.255.0 110
Wait a minute, there are some numbers missing. Where are 127.0.0.0 – 127.255.255.255? IANA (Internet Assigned Numbers Authority) has designated that these are to be used for internal networks only. That having been said, pay close attention to these numbers, as knowing them is the first step to understanding subnetting.
Let’s look at networks and hosts. In the above figure, the numbers indicate the network portion or octets of the IP address, while the “x” represents the host portion or octets. Obviously, the more x’s you have, the greater the number of possible hosts on your network.
Class Network Portion # of Networks Host Portion # of Hosts
Class A 1 number (octet) 126 3 x (octets) 16,777,214
Class B 2 numbers (octets) 16,384 2 x (octets) 65,534
Class C 3 numbers (octets) 2,097,152 1 x (octets) 254
Before any actual subnetting can be done, each octet in an IP must be converted to its binary equivalent. We’ll pick the Host IP address of 196.16.8.3 for our example.
196 . 16 . 8 . 3
11000100 . 00010000 . 00001000 . 00000011
Now let’s look at ANDing. The subnet mask filters out the network bits by using ANDing. ANDing takes the IP address in binary value and uses the binary value of the subnet mask to determine the bits that are left as one or 'active'. This is accomplished by taking the first bit (beginning with octet 1) of the IP address and the first bit of the subnet mask and checking the outcome of the two bits. There are only 4 possible outcomes: 1+1=1, 1+0=0, 0+1=0, and 0+0=0
255 . 255 . 255 . 0
11111111 . 11111111 . 11111111 . 00000000
Now let’s AND them:
11000100 . 00010000 . 00001000 . 00000011
11111111 . 11111111 . 11111111 . 00000000
11000100 . 00010000 . 00001000 . 00000000 = 196.16.8.3
So after ANDing, we have 196.16.8.0 as our Subnet; that means we can assign IP addresses to our Hosts of 196.16.8.0 – 196.16.8.255 right? Well, almost. All subnets need a network address and a broadcast address. The Network address will always be the first address, in this case 196.16.8.0. The Broadcast address is the last address in the range, or 196.16.8.255. That leaves us with the useable Host Ips of 196.16.8.1 – 196.16.8.254.
Think you have the hang of it? Here comes the curve ball. The IP address of 196.16.8.3 with a subnet of 255.255.255.0 is usually written as 196.16.8.3 /24. Why the 24? Remember that each octet of an IP is made up of 8 bits. The default 255.255.255.0 can be thought of as 8 bits + 8 bits + 8 bits + 0 bits = 24 bits. So what happens when you see an address of 196.16.8.3 /27? Where do you get the extra 3 bits from? Easy, you borrow them from the fourth octet’s host portion. Now we have 8 bits + 8 bits + 8 bits + 3 bits = 27 bits; lets see how that translates into binary:
11111111 . 11111111 . 11111111 . 11100000
Converted from binary to decimal:
255.255.255.226
This also means that we are now limited to only 5 bits for our Host portion of the IP. The rule of thumb is that when Host bits go down, the maximum number of subnets goes up, but the amount of possible hosts on each network goes down. Confused? Let’s look at both of our examples together:
196.16.8.3 /24 196.16.8.3 /27
11000100.00010000.00001000.00000011 11000100.00010000.00001000.00000011
Subnet Mask Subnet Mask
11111111.11111111.11111111.00000000 11111111.11111111.11111111.11100000
See where the 1’s stop on our /27 network? That is the end of the Network portion, and the 0’s mark the beginning of the host portion. So, our possible networks have gone up, but the number of possible hosts on each network has decreased. In our original example (/24) we had 1 possible subnet with 253 useable hosts; now with our /27 subnet, we have 8 possible subnets but only 30 possible hosts. It’s a trade off, and depending on the size of your network you can play with the numbers to fit your desired needs.
Terms You Just Ought to Know
