Variable Length Subnet Masking (VLSM) is the best thing to happen to IP networking since subnet masks. In a nutshell, according to what we’ll call classful subnetting rules, if a network is divided into subnets, each subnet within the classful network must be the same size. Figure 36-2 shows the ways in which a Class C network can be divided. Under the traditional classful rules, these are the only divisions possible. The reason for this is the binary nature of the subnet mask.
Figure 36-2. Classful subnets of a /24 network
When IP was developed, there were more IP addresses available than anyone could have possibly wanted—or so we all thought! But as the world progressed, we quickly discovered that we were running out of IP address space at an alarming rate. Some bright minds determined that if the rules of subnetting could be altered a bit, we could give ourselves some breathing room. The result was VLSM.
VLSM states that when a network is subnetted, the individual subnets do not need be of equal size. The caveat is that the subnets must still follow the rules of binary; that is, they must exist where they would normally fall when you’re subnetting using the traditional rules.
Figure 36-3 shows examples of a normally subnetted network, a proper VLSM network, and an invalid VLSM network. In the invalid example, the network 192.168.1.200/28 is not permitted by the rules of subnetting. ...