Basic CIDR question here, given a range of 10.0.0.0/16, and wanting to divide this range into smaller subnets, does CIDR require that you use equal sized subnets, i.e. all of the subnets would need to be /21s or /23s but not both?
For example, is there anything wrong with this combination:
10.0.0.0 / 23 network id 10.0.0.0 broadcast 10.0.1.255
10.0.2.0/ 21 network id 10.0.2.0 broadcast 10.0.9.255
I'd have thought you could start your subnet at any point as defined by the Network ID and broadcast address of the assigned subnet but all of the CIDR calculators I see online do not allow you to provide the net id or broadcast, and seem to assume you are using equal sized subnets all the way through the range.
If I offered the address 10.0.2.0/21 it would assume that the network id was 10.0.0.0 and the broadcast was 10.0.7.255 for this range rather than 10.0.2.0 and 10.0.9.255 so in this instance I'd have to start using my /21 at 10.0.8.0/21 if my /23 had already been provisioned.
Fundamentally then, if I have a /22, I need to take multiples of 4 to find a start point, or a /28 is multiples of 16 to find a start point for the network Id, can anyone advise why this is, as it seems like it could lead to waste by having gaps between subnets?
