IP Address
- 32 bits length
- 4 fields of 8 bits separated by a dot (Dotted Decimal Format)
- Max value of 255 (decimal) per octet
- Classful IP Addressing
- Subnets
- Classless IP Addressing
Classful IP Addressing
Originally, IP addressing was broken down into classes called CLASSFull IP Addressing. Nowadays, IP addresses are broken down classlessly.
| Class | Value of 1st octet | Subnet mask | Classless Interdomain Routing Notation (CIDR) | Example |
|---|---|---|---|---|
| A | 1-126 | 255.0.0.0 | /8 | 10.10.10.10/8 |
| B | 128-191 | 255.255.0.0 | /16 | 150.101.45.45/16 |
| C | 192-223 | 255.255.255.0 | /24 | 200.0.0.30/24 |
| D | 224-239 | N/A | N/A | 227.0.0.30 |
| E | 240-255 | 255.255.0.0 | /16 | 150.101.45.45/16 |
Class D is used for Multicast. It is a one-to-many transmission but the host has to be subscribed to it for receiving data, unlike Broadcast where you receive data no matter if you are subscribed or not.
Class E is Experimental.
127.0.0.0
The whole range 127.0.0.0 is reserved for testing.
127.0.0.1 is called the Local Loopback. If you can reach this address (for example, with a ping), then TCP/IP protocol stack was properly installed on your machine.
Subnet
Your Classful IP Address has a network portion_ and a host portion.
A Subnet separates the host portion of the address from the network portion.
For any bit in the IP address, if the corresponding bit in the subnet mask is on, then that mean it belongs to the network portion of the address.
You can then deduce number of hosts with 2^n - 1 where n is the number of bits in host portion of the address
Must Know
- You cannot use the first address x.x.x.0 and the last address x.x.x.255 in any network for your PC.
- First address x.x.x.0 is for the network.
- Last address x.x.x.255 is the broadcast address.
- Then you can only assign from x.x.x.1 to x.x.x.254
Classless IP Addressing
Classful IP addressing is unsuitable if we want to 'buy' less than 254 addresses, so if a company needs 30 addresses, how do we do that?
Let's see this with an example.
Suppose we have 150.101.45.0/27 in CIDR notatation, we know its submask will be 255.255.255.224. We only needed to look at the last octet to know this, which in binary is 1110 0000 = 128 + 64 + 32 + 0 0 0 0 0 = 224.
The last bit indicates our 'block size', so we have network addresses at 150.101.45.0, 150.101.45.32, 150.101.45.64, 150.101.45.96, 150.101.45.128, all the way to x.x.x.255.
The hosts are the reamining 0s. We also know that the remaining bits for the hosts can only get up to 31 in decimal, also, that the first address 150.101.45.0 is reserved for the Network Address, and the last one 150.101.45.31 is reserved for Broadcast, so we end up with a range of 150.101.45.1 to 150.101.45.30 30 valid addresses.
Then if you get asked to buy 30 addresses, you need to buy an IP addresses with a /27 mask or a 255.255.255.224 mask.