Subnetting - Divide networks into right size

Every network within the internetwork of a corporation or organization is designed to accommodate a finite number of hosts.
Some networks, such as point-to-point WAN links, only require a maximum of two hosts. Other networks, such as a user LAN in a large building or department, may need to accommodate hundreds of hosts. Network administrators need to devise the internetwork addressing scheme to accommodate the maximum number of hosts for each network. The number of hosts in each division should allow for growth in the number of hosts. This is subnetting made easy explanation

Determine the Total Number of Hosts

First, consider the total number of hosts required by the entire corporate internetwork. We must use a block of addresses that is large enough to accommodate all devices in all the corporate networks. This includes end user devices, servers, intermediate devices, and router interfaces.

See Step 1 of the figure.

Consider the example of a corporate internetwork that needs to accommodate 800 hosts in its four locations.

Determine the Number and Size of the Networks

Next, consider the number of networks and the size of each required based on common groupings of hosts.

See Step 2 of the figure.

We subnet our network to overcome issues with location, size, and control. In designing the addressing, we consider the factors for grouping the hosts that we discussed previously:
Grouping based on common geographic location
Grouping hosts used for specific purposes
Grouping based on ownership

Each WAN link is a network. We create subnets for the WAN that interconnect different geographic locations. When connecting the different locations, we use a router to account for the hardware differences between the LANs and the WAN.

Although hosts in a common geographic location typically comprise a single block of addresses, we may need to subnet this block to form additional networks at each location. We need to create subnetworks at the different locations that have hosts for common user needs. We may also have other groups of users that require many network resources, or we may have many users that require their own subnetwork. Additionally, we may have subnetworks for special hosts such as servers. Each of these factors needs to be considered in the network count.

We also have to consider any special security or administrative ownership needs that require additional networks.

One useful tool in this address planning process is a network diagram. A diagram allows us to see the networks and make a more accurate count.

To accommodate 800 hosts in the company's four locations, we use binary arithmetic to allocate a /22 block (2^10-2=1022).

Allocating Addresses

Now that we have a count of the networks and the number of hosts for each network, we need to start allocating addresses from our overall block of addresses.

See Step 3 of the figure.

This process begins by allocating network addresses for locations of special networks. We start with the locations that require the most hosts and work down to the point-to-point links. This process ensures that large enough blocks of addresses are made available to accommodate the hosts and networks for these locations.

When making Subnetting which divides networks into right size and assignment of available subnets, make sure that there are adequately-sized address blocks available for the larger demands. Also, plan carefully to ensure that the address blocks assigned to the subnet do not overlap.