IP Address Subnetting is a complicated topic that can be confusing for many students. While you may be using one of the free subnet calculators available on the internet, knowing the reasons behind it and why it works the way it does is critical for true success. This white paper explains the basics of IP version 4 addressing and subnetting.
The specifications for IP version 4 were published in 1981 and remain in use today. The next version will be version 6. This document explains the basics of IP version 4 addressing.
IP is an OSI Reference Model Layer 3 protocol; its primary function is to deliver datagrams to the destination host based on the logical, network manager-assigned, host IP address.
The destination host may be on the same network as the source or on a different network. Before transmission, the source host must determine if the destination is local or remote. The IP protocol needs a facility to make this determination.
One way is to classify addresses. The classes would determine what portion of the IP address is common to all hosts on the same network. This allows a transmitting host to compare its address to the destination. If the network portion of the source IP address and the destination address are the same, the destination is local and the datagram can be delivered directly. Otherwise it must be delivered indirectly by forwarding the datagram to a router that will then forward it towards the destination.
An IP address is made up of 32 bits. While computers have an easy time dealing with a sequence of 32 binary digits, 1s and 0s, most human beings aren't quite as comfortable so addresses are divided into four groups of eight. Each group of eight is referred to as an octet and an address is usually represented as a sequence of four decimal byte values separated by dots. 10.50.60.21 is an example.
The class of an address can be determined by inspecting the first byte of the address. The following table shows the different classes.
From the table above we can determine that our example, 10.50.60.21, is a Class A address. It belongs to a device on network 10, and 50.60.21 is the unique host portion. By inspecting its own address, this device determines that it can directly contact any other host with 10 as the byte value of the first octet of its address. All other destinations must be routed through a router. Similarly, hosts with Class B addresses inspect the first two octets, Class C the first three.
While easy to understand and implement, this plan proves inefficient to deal with larger networks. Consider that using this technique, a Class A network can only be segmented based on Data Link Layer addressing. Since Data Link addressing identifies each host individually without providing a "network" portion to the address, the only devices that can be used to segment such networks are bridges or switches. These devices must maintain a bridging table with an entry for every host on the network. In a Class A network, each bridge could have up to 16,777,214 entries in its table. Moving data from network segment to network segment takes a long time since the bridge must look up the destination address in a very large table. Bridging proves insufficient in large networks, and often we can't bridge between different technologies such as LAN and WAN. In these cases, it is much preferred to use routers since they forward packets in large networks more efficiently and can function with mixed technologies.
Imagine the following scenario. A large organization has many facilities in different cities and many departments. The network administrators want an easy way to assign IP addresses that allows them to easily identify the computer, department, and city. Using Class A network address 10.0.0.0, and representing the facility in Montreal with the number 50, 60 representing the Sales department. The IP address of 10.50.60.21 references computer 21 in the sales department in Montreal.
For human beings this is great, but for IP hosts this could be a problem. Imagine that this computer wants to send a packet to computer 33 in the marketing department in New York City, 10.100.70.33. Computer 10.50.60.21 must determine if the destination is on the same LAN or accessible via a router. By examining its own IP address, the source host determines that all hosts whose addresses begin with the value 10 are on the same LAN and can be contacted without using a router. Since the facilities in Montreal and New York are linked by a WAN, and all traffic between the facilities must be forwarded by routers, the transmission will fail.
No comments:
Post a Comment