After reading this article you will learn about the definition of subnets.
All hosts in a network must have the same network number—a fact that must have become obvious from the above discussion on IP addressing. As networks grow, this may cause problems. For example, consider a newly established company. Naturally, when it starts, it may have just one class C LAN on the Internet.
However, as it grows, it may acquire more than 254 computers in its LAN and, therefore, need another class C address. It may also acquire a different LAN and might need a new IP address for this new LAN. The LANs could be bridged to form a single IP network, but these bridges could create their own set of problems. This continued growth could result in several separate LANs, each with its own IP network address.
As the number of such local networks grows, keeping track of them and managing them can become quite a serious problem. For example, every time a new network is installed, the system administrator will have to get in touch with NIC and get a new network number. This number must be advertised world-wide.
ADVERTISEMENTS:
Moreover, if a machine has to be changed from one LAN to another, its IP address will have to be changed. This in turn may require modifications of its configuration files. In short, this growth of an existing LAN will cause a series of very large headaches.
In order to obviate these problems, if the network is allowed to be split into several parts for internal use but acts like a single network for the outside world, then all the headaches can be avoided. These parts are called subnets. If the subject company that we are discussing started up with a class B address instead of a class C address then it could start out by just numbering the hosts from 1 to 254.
When the second LAN is added, it could split the 16-bit heist number into a 6-bit subnet number along with a 10- bit host number. This split could allow 62 LANs, each with up to 1022 hosts. This-scheme is illustrated in Fig. 10.3.
Outside of the network, the fact that there is a subnet in operation is not visible. Therefore, allocation of the subnet will become a totally internal matter and no reference need be made to NIC. Being a class B network, the IP addresses for the first subnet could, for example, start from 140.60.4.1 and for the second subnet from 140.60.8.1.
ADVERTISEMENTS:
To understand the subnet operation, it is necessary to understand how IP packets are processed at a router. Each router has a table which has some listings of IP addresses. The first kind tells how to get to distant addresses. The second tells how to get to local hosts. Associated with each table is the network interface to be used to reach the destination.
When the IP packet arrives, its destination address is looked up in the routing table and if the packet is destined for a distant network, it is forwarded to the next router on the interface given in the table. If it is a local host, it is sent directly to the destination.
If the network is not present in the table the packet is forwarded to a default router which has more extensive tables. This implies that each router has to only has to keep track of other networks and local hosts, not network-host pairs. This greatly reduces the size of the routing table.