Network class
Network classes ( Classful Network ) were a division of the IPv4 address range in the Internet Protocol into subnetworks for different users, used from 1981 to 1993 . The size of a network could be derived from the network class, ie with IPv4 the number of bits for the network portion in the IP address. This is important for routing in the intranet and Internet in order to distinguish whether a destination IP address can be found in your own network or in a foreign network. Since network classes have proven to be too inflexible and not very economical in dealing with the scarce resource IP addresses, they were initially supplemented in 1985 with subnetting and in 1992 with supernetting and finally in 1993 with the introduction of Classless Inter-Domain Routing ('CIDR' for short) ) replaced. Regardless of this, the outdated and no longer practice-relevant concept of network classes is often still taught by lecturers and can often be found in textbooks.
Concept of net classes
The originally used concept of IP addresses only provided for a rigid division. 8 bits were provided for addressing the network, the remaining 24 bits addressed a specific participant in the network. With this concept, however, only 256 networks were possible. This was recognized as too little. Therefore, in September 1981, the so-called network classes were introduced by RFC 791 , which redesigned this division.
The entire address space was initially divided into three (later five) network classes via the network classes. All sub-networks of a network class had the same standardized size. The network sizes of the classes differed greatly; in a class C network only 254 hosts were possible, whereas in a class A network over 16 million hosts were possible. This should make it possible to assign networks of different sizes to individual organizations and institutions as required. But the rigid network sizes led to great waste, since z. For example, a user with 100,000 hosts had to be assigned a class A network. Of these, however, only 126 were available and in this specific case over 16 million IP addresses would have been wasted. Therefore, the IP classes were replaced in 1993 by RFC 1518 and RFC 1519 by classless inter-domain routing . With CIDR , networks of flexible sizes are assigned within the entire address space, so it is no longer possible to derive the network size from the IP address.
The network class was determined by the first bits of the binary IP address. The area assigned to classes D and E was reserved for extended addressing in the original specification. This was later divided into classes D and E, with the address range of class D still being used for multicast applications even after the network classes were abolished . The address range of the former class E is still reserved.
Overview of the network classes
| Network class | prefix | Address range | Netmask | Net length (with prefix) |
Net length (without prefix) |
Host length | Networks | Hosts per network |
CIDR suffix equivalent |
|---|---|---|---|---|---|---|---|---|---|
| Class a | 0 ... | 0.0.0.0 - 127.255.255.255 | 255.0.0.0 | 8 bit | 7 bit | 24 bit | 128 | 16,777,214 | /8th |
| class B | 10 ... | 128.0.0.0 - 191.255.255.255 | 255.255.0.0 | 16 bit | 14 bit | 16 bit | 16,384 | 65,534 | / 16 |
| Class C | 110 ... | 192.0.0.0 - 223.255.255.255 | 255.255.255.0 | 24 bit | 21 bit | 8 bit | 2,097,152 | 254 | / 24 |
| Class D. | 1110 ... | 224.0.0.0 - 239.255.255.255 | Use for multicast applications | ||||||
| Class E. | 1111 ... | 240.0.0.0 - 255.255.255.255 | reserved (for future use) | ||||||
Outdated teaching
However, this doctrine of network classes often only leads to confusion, as it is obsolete with the introduction of CIDR . Today it has almost no practical relevance, since the size of a network can no longer be deduced from the IP address, but the specification of a network mask is mandatory . The technology behind it is called VLSM .
Even today, however, some implementations still follow the old network class concept: The point-to-point protocol mechanism under Windows (CE, XP) derives the size of the network for a direct connection between two PCs from the IP address; a network mask is not entered intended. The network class is still of practical importance when using historical routing protocols such as B. RIPv1 . Since these older routing protocols, which are still occasionally used in LANs, do not transmit a network mask, the network mask belonging to the network class of the IP address is automatically adopted if necessary. The same applies to IP implementations in historical operating systems that have not been further developed since the standardization of subnetting in 1985. The historical network classes are also the background of the reserved address spaces for private IP addresses .
The diminishing importance of network classes is also reflected in the IP allocation policy of the Regional Internet Registries (RIR). In the former class C area, there are also allocations that go far beyond the size of a network of the old class C (the associated mechanism is supernetting ). Similarly, due to a shortage of IP addresses, the former class A areas are now assigned in smaller blocks, such as the area 80.128.0.0 to 80.159.255.255 taken from the former class A network of number 80 to Deutsche Telekom AG .
An overview of the network sizes currently and in the past assigned in the RIPE region can be found in document RIPE-345 on the RIPE NCC website .
See also
Individual evidence
- ↑ Smallest RIPE NCC Allocation / Assignment Sizes ( English ) ripe.net. Archived from the original on February 9, 2018. Retrieved September 13, 2019.