The IP address of the sender and receiver are included in each data packet, and the only information network devices have regarding the package's destination. Domain name is not specified in the individual packages. Each device can be listed as the addressee on the Internet must have its own IP address. This applies, for example, routers, infrastructure servers (for example, NTP, DNS, DHCP), common personal and server computers, network printers, and independent IP phones. With special arrangements computers can share the address, but they are seen as the basis of a computer.
Devices that are not to be seen or be directly addressed other hand, need no address. Examples of such devices are switches, bridges and firewalls. They can still have their own addresses given remote administration or other functions integrated in the same device.
Originally addresses are intended to be unique on the Internet, but some address blocks have since been allocated for use in individual intranet. Some addresses or address blocks are also intended for special purposes where they relate to a single appliance, such as network address and broadcast address for an individual smaller networks (subnets), multicast address for those interested in a particular service
(one TV channel or the like) and Anycast addresses with which to reach an arbitrary server that provides a particular service.
Users on the Internet rarely use IP addresses directly. Instead, the domain name by the domain name system (DNS), where the computers are named in a hierarchy according to the organization and organization type or country.
When a computer should be contacted takes first track of the name server for the organization responsible for the name, for and by this get the IP address of the computer in question. Then use this IP address. In some protocols, the name may have significance later, but it is not used to control the data packets.
The relationship between the domain name and IP address do not need to be unambiguous and the IP address works independently of the DNS. Conversely, those who registered a domain name using this and administer a name server does not hold any IP address blocks. If the name and IP address do not match
(When controlling the other way), this can be taken as a sign that all is not right, the organization responsible for the name and the person responsible for the address should generally cooperate.
IP addresses are handed out en bloc, so that by analyzing the first part of the address can infer which network it belongs. Thus, they need central routers on the Internet do not know all the specific IP addresses, but only the router that is responsible for specific blocks. Generally, an IP address is reached different paths and advanced arrangements are to the core network routers, and routers that have multiple routes to the main grid, to take appropriate decisions on which path a given packet should be sent.
Every computer connected to the Internet, or to another Internet, an IP address, sometimes more, the network administrator or Internet service provider, in any nätblock that the organization awarded
(Or in one block for local addresses).
The address can either be configured by hand on the individual computer or assigned dynamically through DHCP or similar arrangement in connection with the computer is started up. IP address changes occasionally, either because of the reorganization of the network or, for dynamic allocation, then the computer one time were not connected and address therefore assigned to another machine. Often recorded task on any computer or connection at a certain time used a particular IP address
(Legislation in the EU gives ISPs the obligation to post the use).
Unable to safely bind a specific IP address to a particular person, then it need not be the owner of subscriptions that use the subscription. Moreover, it is now common to consciously distribute their bandwidth with such FON and Tor, or unconsciously, by neglecting to secure their wireless networks.
Some IP address blocks used for special purposes. So, for example, block 127.0.0.1/8 in IP version
4 reserved for use within a single computer ("localhost"), 0.0.0.0/8 for the local nätsegmentet, 169.254.0.0/16 for automatically configured local communication ("linklocal"), 192.0.2.0/24 for example (cf. example.com) and 18.104.22.168/4 Multicast.
IP version 4 (IPv4) is the version of IP used since the Internet's breakthrough. In IPv4 comprises a
IP number of 32 bits, which is equivalent to 4 bytes and allows the number of addresses is limited to just over four billion (4,294,967,296 or 232).
When IPv4 was developed, it was mainly multi-user, server computers and networking equipment that needed to addresses, and addresses appeared enough with very large margin. As personal computers become common, and mobile phones and soon also consumer electronics connected to the Internet holds addresses running out. Since the 1990s it has been warned of the problem and IP version 6 was developed primarily to increase the amount of available addresses, but has not yet been seen.
Usually the IP address by printing out the 4 bytes in decimal form with dots in between, so that each digit group represents one byte.
Examples of IPv4 addresses:
127.0.0.1 (the internal network address of a device, also called loopback)
Some IP addresses are used for local area networks (LANs) in homes and offices. These addresses are reserved by the Internet Assigned Numbers Authority (IANA) and can not be used directly on the Internet. For computers with such local addresses should be able to connect with the Internet requires a router with Network Address Translation (NAT), ie a router that replaces the addresses of individual data packets to those once only on the Internet, and to addresses on the intranet.
The local addresses are (defined in RFC 1918, previously divided into A-, B- and C-class networks):
10.0.0.0 - 10255255255
172.16.0.0 - 172.31.255.255
192.168.0.0 - 192.168.255.255
The new standard of IP called IP version 6 (IPv6). In IPv6 IP number increased to 128 bits. Also, several new features have been introduced, such as IPSec (which later also became ported to IPv4). The notation for addresses are also clearly different, with a format of a total of 128 bits arranged in eight groups of hexadecimal digits and colons (:) as a separator between the groups.
A subnet (LAN) is assigned a number of IP addresses in sequence and form that way a network. Where the boundary between LAN (address fields) is defined by a subnet mask. Nowadays the mask only specifies the number of bits in the network address and the local part of the address, previously more complex configurations allowed.
The network addresses are only some bits meaningful when the individual addresses refers to computers or devices on that network. The web address listed so that the other pieces count as zero and one for the address adds a slash and either the number of meaningful bits at the beginning of the address, or a "netmask" where the meaningful bits replaced with ones and the insignificant zeros. So, for example 192.168.0.0/24 192.168.3.0/24 and different network addresses by 255 each, and 192.168.0.0. / 255.255.255.0 is an alternative way to describe the first-mentioned network. The address where the network address meaningless bits is the "broadcast address" of the network in question, and refers to all online (such addresses often blocked by firewalls to prevent network overload by the response to malicious requests to the address).
To addresses in IPv4, suffice it requires that certain allocated dynamically. The addresses can thus be used by multiple customers, but not simultaneously. So-called fixed IP number assigned primarily computers or local networks of particular causes need such a, such as server computers. Often, companies and other organizations fixed addresses on their computers, then it can ease administration. Home users often have dynamic IP number when computers are often muted and dynamic addresses can thus be used more effectively: the number of addresses corresponding to the maximum number of simultaneously connected enough, regardless of the number of customers. A recommendation on how addresses are allocated among internet providers available in the RFC in 2050.
A computer can have a permanent address, as hand written into the appropriate configuration files, or ask at the start after his address and get this information from a BOOTP or DHCP server. If the computer is firmly connected to the network can get the address when connecting to a new network, for example via a modem pool (PPP), a 3Gmodem or a common wireless LAN (WLAN). The server can be set to always give the same fixed address access certain computers while others computers share a limited number of dynamic addresses. Usually, however, strive server after not exchange addresses unnecessarily.
Previously, IP addresses are divided into classes, Class A networks had 224 addresses, network Class B 216 addresses and network of class C 28 addresses. This division was abolished in the 1990s to the addresses could be allocated more efficiently.