Illustrated TCP/IP by Matthew G. Naugle Wiley Computer Publishing, John Wiley & Sons, Inc. ISBN: 0471196568   Pub Date: 11/01/98

Chapter 79
Supernetting Exposed

In the previous example, we showed a subnet mask for a Class C address that was shorter than the natural mask. Applying this to the example, the ISP has a block of addresses. As far as the ISP is concerned, there is no Class associated with the address; it is simply a block of addresses defined by the prefix. This block is assigned to the ISP by the Internet Registry under the authority of the Internet Assigned Numbers Authority (IANA, yes, the same group, actually only one person who handles the top-level domains). This is small (four Class C addresses), but it shows up as one entry in the routing table 200.1.252.0/22. Notice the mask at the ISP is pushed back to the left beyond the natural subnet mask of a Class address. This is known as supernetting.

The current approach (in lieu of IPv6) is to provide large contiguous blocks of Class C (and possibly other classes) addresses. They are provided by more local levels in an hierarchical fashion. For example, a national backbone provider (call it ISP-1) with connections to other national backbone providers through Network Access Points (NAPs) will be assigned a large block (one that will last two years) of Class C addresses. In turn, other regional service providers (call them ISP-2) who utilize ISP-1 will be assigned a block of addresses from ISP-1’s address block assignment. In turn, ISP-2 will provide address assignment to its customers from the block it was assigned. This allows for very efficient and manageable global routing tables (those routing tables on the top-level providers).

Supernetting Exposed