What Is a Supernet?

Supernetting is a similar concept to subnets-IP addresses are divided into independent network addresses and host addresses based on the subnet mask. Supernetting, also known as Classless Inter-Domain Routing (CIDR), is a method of aggregating multiple similar Internet addresses.

Supernetting is a similar concept to subnets-IP addresses are divided into independent network addresses and host addresses based on the subnet mask. Supernetting, also known as Classless Inter-Domain Routing (CIDR), is a method of aggregating multiple similar Internet addresses.

Chinese name

Supernet

Foreign name

supernetting

Definition

Is a similar concept to subnets-IP addresses

Advantages

Can make full use of C-type cyberspace resources

Supernet concept

Supernetting is a similar concept to subnets-IP addresses are divided into independent network addresses and host addresses based on the subnet mask. However, in contrast to the subnet dividing a large network into several small networks, it is a combination of some small networks into a large network-the supernet.

Supernet was created to solve the problem that the routing list exceeds the existing software and management manpower, and to provide a solution for the exhaustion of class B network address space. Supernet allows a route list entry to represent a collection of networks, just as an area code represents a collection of telephone numbers for an area.

The current popular external gateway protocols Border Gateway Protocol (BGP) and Open Shortest Path First (OSPF) routing protocols support supernet technology. ^[1]

Subnetting divides a single IP address into multiple subnets to delay the allocation of large network addresses (mainly Class B) ^[2] . Subnetting has indeed played this role since it was proposed in the 1980s. However, in the 1990s, subnetting could not prevent the tendency of Class B network addresses to run out. The reason is simple. There are only over 16,000 Class B addresses. When applying for an address for a medium-sized network, people are more inclined to use class B addresses and subnetting them to avoid inconvenience caused by the use of multiple class C addresses for network configuration and management. Therefore, the allocation of Class B addresses is fast, while the allocation of Class C addresses is much slower. In order to solve the problem of class B address space tension and make full use of the class C address space (the number of class C networks is more than 2 million), people have also proposed supernet technology.

The function of Supernet is to aggregate multiple consecutive Class C network addresses and map them onto a physical network. In this way, the physical network can use the common address prefix of the aggregated class C address as its network number.

Supernet was created to solve the problem that the routing list exceeds existing software and management manpower, and to provide a solution to the exhaustion of class B network address space. Supernet allows a route list entry to represent a collection of networks, just as an area code represents a collection of telephone numbers for an area.

The current popular external gateway protocols Border Gateway Protocol (BGP) and Open Shortest Path First (OSPF) routing protocols support supernet technology. ^[1]

Supernet advantages and disadvantages

The advantage of supernet is that it can make full use of C-type network space resources ^[2] . In most cases, the use of supernet address allocation is willing to match the allocated network space with the actual number of nodes required, thereby improving the utilization of the address space. For example, for a 4000-node physical network, assigning a class B address is obviously wasteful, but the class C address is too small, so we can assign an address space block consisting of 16 consecutive class C networks to the physical network.

The super-network method also brings a new problem: the growth of the routing table size. The size of the routing table is directly proportional to the number of networks. A physical network corresponds to multiple class C network addresses, making the network correspond to multiple class C prefix entries in the routing table, making the routing table too large. The overhead brought by routing protocols to exchange routing information has also increased dramatically. This problem can be solved by using Class Inter-Domain-Routing (CIDR) technology. Although a physical network corresponds to multiple entries in the routing table, all entries must point to the same next hop address, so it is possible to aggregate entries. CIDR technology can aggregate consecutive class C network address blocks in the routing table. The class C network addresses must be continuous, and the number of address blocks is a power of two. The length of the network prefix of the aggregated CIDR address block. Obviously, the length of the subnet mask will be less than 24 (the mask length of the class C network). Like the representation used in subnet routing, CIDR defines fast addresses and is uniformly expressed in the form of "network prefix / number of subnet mask bits".

CIDR Features of Supernet CIDR

(1) CIDR eliminates the traditional concepts of class A, B, and C addresses and the division of subnets, and can more effectively allocate IP address space. CIDR uses "network prefixes" of various lengths instead of network numbers and subnet numbers in classified addresses, instead of using only 1-byte, 2-byte, and 3-byte network numbers in classified addresses. CIDR no longer uses the concept of "subnets" but uses network prefixes, and uses IP addresses to return from three-level addressing to two-level addressing, that is, two-level addressing without classification. ^[2]

IP address :: = {<network prefix>, <host number>}
CIDR also uses "slash notation", that is, write a slash "/" after the IP address, and then write the number of bits occupied by the network prefix (corresponding to the number of 1 in the subnet mask).

(2) CIDR composes continuous IP addresses with the same network prefix into a "CIDR address block". A CIDR address block consists of the start address of the address block (that is, the one with the smallest address value in the address block) and the address in the address block. To define. CIDR address blocks can also be represented by slash notation. ^[2]

Since a CIDR address block can represent many addresses, the CIDR address block is used in the routing table to find the destination network. This type of address aggregation is usually called route aggregation, which allows an entry in the routing table to represent many routes of the original traditional classified address. Route aggregation is also known as forming a supernet. Route aggregation helps reduce the exchange of routing information between routers, thereby improving the performance of the entire Internet.

Examples of supernet applications

A physical network of 4,000 nodes is connected to the Internet through router R, and is assigned an address space composed of 16 consecutive class C network addresses ^[2] . In order to route this network as a unified network, on router R, CIDR can be used to aggregate the routing entries of the 16 class C addresses into one entry for routing. Each router in the network only needs to record a routing entry of a network address, instead of recording a separate entry for each of the ^[3] class C network addresses.

This type of continuous Class C address block represented by the symbol "network prefix / subnet mask" is called a CIDR block. As a method to reduce the speed of IP address allocation and reduce the number of entries in Internet routers, CIDR technology has been widely recognized and widely used in address allocation. When applying for an Internet address, the Internet address obtained from the address assignment agency is often a CIDR block, rather than a traditional classified address.

In fact, CIDR does not restrict the network address to be class C. Therefore, CIDR does provide a "typeless" address allocation method. In terms of routing table representation and routing, CIDR and subnets are similar.

The IETF reserves a part of the CIDR block in the entire IP address space for a private network, called a private address. These addresses are not assigned to networks and nodes on the Internet by Internet address management agencies. Private addresses are not legal addresses in the Internet, so the address space of the private network is isolated from the Internet address space. This isolation provides a security means to protect the internal network from intrusions from the fabric. Of course, isolation also prevents the internal nodes of the private network from directly accessing the Internet, and requires the use of a proxy or NAT (network address translation) mechanism at the network entrance.

When doing ACL, in order to save equipment costs (that is, use as few ACL entries as possible), supernet technology is used. Therefore, the calculation and division of the super network are involved. Here are some examples to illustrate the method of calculating the supernet:

1. It consists of 2048 Class C networks, from 192.24.0.0 to 192.31.255.0. Which mask can represent this address range?

2. When I configured the ACL, I encountered such an address range: 59.39.80.0-59.39.127.0, how to divide the supernet to minimize the ACL entries?

Reference solution:

1. 192.24.0.0 is divided into binary as

11000000.00011 000.00000000.00000000

192.31.0.0 is divided into binary as

11000000.00011 111.00000000.00000000

-identical

Take the same number of bits from left to right as the number of subnet masks, that is: 13 bits, the mask is 255.248.0.0

2. 59.39.80.0 converted to binary is:

00111011.00100111.01010000.00000000

Take the digits before all 0 as the network address bits, that is, 59.39.80.0/20 [59.39.80.0-59.39.95.255] to get a supernet

Then the remaining addresses start with 59.39.96.0,

59.39.96.0 converted to binary is:

00111011.00100111.01100000.00000000

The digits before all zeros are the network address bits, that is, 59.39.96.0/19[59.39.96.0-59.39.127.255], all of which cover this address pool.

A total of two ACLs were obtained, respectively: 59.39.80.0/20, 59.39.96.0/19