In truth, the subnet does have to conform to the rule as described. With VLSM, however, it is often forgotten that the entire subnet area is considered one subnet. Therefore, the rule must be obeyed once, not on each instance of variable subnetting.
In the preceding example, you would choose to obey the rule either in the bits allocated to the region, campus, or building, but not in each hierarchical layer. It would make most sense to adhere to the rule using the least-significant bits. In this case, three bits have been allocated to the access layer, enabling you to identify eight floors. You have no more than three floors to address in any building, however. Obeying the rule on this layer makes sense because you reduce the floors that may be addressed to six, which is still twice as many as required.
Was this article helpful?