Do you mean to have more than one IP prefix to subnet A? Couple of questions comes to my head for such design:
- how will you uniquely identify subnet A in this case? As you can no longer take the longest prefix match of all the IPs in A as there will be overlap with IPs from other subnets.
- Remember a subnet needs a network address and a broadcast address, what would be they be?
- How do you know when to stop the sub-division? someone might exploit this and create multiple subnets of /31.
- Also wouldn't this add complexity to how you maintain the routing table? Recall that we would like to avoid having big routing tables so we try to merge IPs that could be grouped together in the same row.
- how will you uniquely identify subnet A in this case? As you can no longer take the longest prefix match of all the IPs in A as there will be overlap with IPs from other subnets.
- Remember a subnet needs a network address and a broadcast address, what would be they be?
- How do you know when to stop the sub-division? someone might exploit this and create multiple subnets of /31.
- Also wouldn't this add complexity to how you maintain the routing table? Recall that we would like to avoid having big routing tables so we try to merge IPs that could be grouped together in the same row.
