Supernet vs. Subnet

Before any discussion of supernets or subnets can happen, you have to first understand the “mask” of an IP address. As part of the standard of Classless Interdomain Routing, or CIDR, IP addresses can be delivered with a corresponding subnet mask. Both the IP address and subnet mask are comprised of four octets of binary positions, separated by decimal points. The individual positions of each octet represent, in order from left to right, the values 128, 64, 32, 16, 8, 4, 2 and 1. The highest decimal value of a single octet is 255. Therefore the decimal IP address of 192.168.1.1 is represented, in binary, as 11000000.10101000.00000001.00000001.

While an IP address can be virtually any number, subnet masks are defined by which positions, from left to right, represent the “network” number, and which bits, from right to left, represent the “hosts.” All of the 1’s in a subnet mask represent the network number, while all of the 0’s represent the hosts. There can be no break in the network designation. In the above IP address example, if you wanted to label the “192.168” portion as the network and the “.1.1” portion as the host, then you would need a subnet mask of 11111111.11111111.00000000.00000000, or 255.255.0.0 in decimal. This is sometimes also written as 192.168.1.1/16, with the “/16” meaning that 16 bits are reserved for the network address.

When planning your networks, you need to look at how many routable networks are needed versus how many hosts are expected in each network. In the above example, the hosts 192.168.1.1 and 192.168.1.135 would be on the same network. If you changed the subnet to 11111111.11111111.11111111.10000000, or 255.255.255.128, then the two hosts are then on two different networks -- one on the 192.168.1.0 network, and the other on the 192.168.1.128 network. Each of these networks, then, can accommodate 126 host machines, as two host addresses are reserved for the network number and broadcast address. Expanding the subnet to accommodate more networks is the logical process of subnetting -- you get more networks, but fewer hosts per network.

The opposite of this is supernetting. Supernetting looks at the number of subnets and seeks to implement a mask which encompasses all of those subnets. Therefore, if you wish to combine the two aforementioned subnets into a single supernet, you would simply eliminate one of the network bits. The subnet masks are 255.255.255.128, while the supernet mask are 255.255.255.0, or 11111111.11111111.11111111.00000000.

In simple routing environments, such as a home network, a single private network merely routes to an Internet connection. In complicate routing environments, such as a campus area network that connects multiple buildings and organizations, you must use multiple routers. Subnets help divide and manage users at the local level. For example, if you have graphics design staff and music staff, housed on separate floors of the same humanities building of a university campus, you could use the 192.168.1.0/25 and 192.168.1.128/25 networks to differentiate network traffic. However, when the router at the center of the university needs to “summarize” how to deliver information to the humanities building’s router, it would use 192.168.1.0/24. This allows the central router to simplify its routing and security rules.

Supernetting is also used to break beyond the traditional “class” of a given private network number. For example, 192.168.1.0/24 encompasses hosts from 192.168.1.1 to 192.168.1.254. If you need more than 254 hosts, but wish to retain the “192.168.” structure of the addresses, you can implement a /23 subnet, or 255.255.254.0. This changes the address range from 192.168.1.1 to 192.168.2.254, effectively doubling the number of hosts.