DNS Tutorial Part 1: Naming the Internet.

(From Network Computing Asian Edition)

Byline: Graeme K. Le Roux

The Domain Name Service (DNS) exists for two deceptively simple reasons: to allow humans to deal with text names rather than IP addresses, and to allow one host in one domain to find a host in another domain which provides a specific service, such as a mail server. In the early days of the Internet, the Network Information Centre (NIC)?then the only one?sent a text file called ?hosts? that contains a list of all registered hosts and their corresponding IP to all host administrators via ftp.

Hosts files still exist on most systems which run IP (usually in a path such as \etc\hosts), and in most implementations the hosts file is checked before a call to a DNS server is made. However, the Internet is now simply too large and too dynamic for it to work as a general method for mapping host names and services to addresses.

Web family tree

The DNS is actually the server application for a distributed database, which is referred to as the name space. Within the name space are a number of sub-domains arranged in a tree-like hierarchy. The top of this tree is the null domain. Each DNS server has a section of the global name space called a sub- domain, for which it is responsible.

The first level of sub-domains include national name spaces such as .au for Australia, .sg for Singapore, .uk for the United Kingdom, .jp for Japan, etc., and the original US domains of .com (originally US commercial organisations), .gov (US government), .mil (US military), .edu (US educational institutions), .net (originally main Internet nodes, ISPs, etc.), .org (US non- profit organisations, charities, etc.).