What is an IP number / IP address?
'IP' stands for Internet Protocol - it decribes the way information is sent and received
over the Internet. An IP number, or IP address, is a 32 bit number assigned to any
device that wishes to communicate over the Internet. Typically the IP address is represented as a
set of four numbers in the range 0-255 - the address is usually written as these four numbers
separated by dots. It can also be shown as one big number in decimal or hexadecimal.
What is my IP number?
The IP number of this machine is:
This number can also be represented as 387372190 (32 bit decimal number)
or 0x1716D49E (32 bit hexadecimal number).
(Note that if you are part of an internal network then this is the address of your local server, the
machine which is connected to the external internet.)
How are IP addresses used?
Whenever data is to be transmitted to a particular machine, it is broken up into chunks, or
packets, each of which is tagged with the IP address of the destination machine. Each packet
is transmitted separately, and will not necessarily follow the same route through the network
as the other packets that make up the whole message. It is quite possible that the packets will
arrive out of sequence, or with errors. Some may not even arrive at all.
On receipt the packets are automatically reassembled into the correct sequence so as to
reconstruct the original data;
if there are any errors or missing packets then they are requested to be sent again.
How are IP Addresses Assigned?
IP addresses are hierarchical in nature - that is to say, one part of the address will specify
broadly in which region of the network the destination can be found, with subsequent parts of the
address providing more specific information about the location of the target device within that
region. Consequently the allocation of addresses needs to be carefully managed, in order to
maintain this hierarchy.
The allocation is overseen by the Internet Assigned Numbers Authority (IANA) and Regional
Internet Registries (RIRs), who maintain a publicly accessible database called WHOIS, relating
IP addresses to locations in the network.
What is DNS?
Most Internet users prefer to use addresses which are more meaningful than a sequence of numbers -
for example 'www.microsoft.com' instead of '188.8.131.52'. The translation from the textual
domain name to the numeric IP address is performed by certain machines on the internet, known
as 'Domain Name Servers'. Every computer that wants to be able to use domain names needs to know
the IP address of at least one Domain Name Server - it contacts this server any time it needs to
translate a domain name to an IP address. The server can either reply immediately with the IP address
if it knows it, or it can forward the request to another server which may be in a better position
to provide the information. When the translation is successfully performed, the server will
keep a copy of, or cache, the address so that it can answer the request more rapidly the next time it is asked.
Domain names are resolved starting with the final part, the Top Level Domain (e.g. '.com', or '.de').
Every Domain Name Server knows the locations of certain machines known as Root Servers,
which are able to indicate where to find a machine that can handle the particular Top Level Domain.
The remainder of the domain name is then sent to this machine, which knows where to locate servers that
can deal with the next level in the domain name, and so on until the first part of the name is reached
(e.g. 'www'). At this point a full IP address can be returned through the various servers, back to the
machine which originated the translation request.
Each machine along the return journey will keep a cached copy of the IP address and domain name, so that if
asked for the same address in the near future it can respond immediately without having to contact
any other machines. This speeds up general operation of the network, although it does mean that if
the IP address for a particular domain name is changed at any time, it can take a while -
sometimes several days - for that change
to be propagated through the network. Until the change is propagated, the name servers will continue to
report the old (cached) IP address.
What is IPv6?
An IP address as described above is made up of four bytes, in a format known as IPv4. Four bytes can
be used to represent over four billion different individual addresses, which might seem sufficient to
uniquely label every computer in the world - but with more and more devices being connected to the internet
the number of IPv4 addresses which are free to be allocated is dwindling.
IPv6 mitigates this problem by using sixteen bytes instead of four, which allows approximately 3.4x1038 unique addresses
to be represented. In addition, the standard solves various other technical problems that IPv4 suffers from.
Unfortunately IPv6 is not interoperable with IPv4, so in order to use it all the internet hardware between IPv6
nodes will need to be updated to be able to use IPv6. Until that happens an IPv6 network is effectively
invisible to IPv4 systems, and vice versa, although it is technically possible to implement converter nodes that embed one protocol
within the other to allow connections between the different networks.