Connecting a Client to a Server in Java

Imagine a chat program on your computer that needs to exchange messages with a server running on another machine across the Internet. Before a single message can move, the program has to answer a question that sounds simple: which computer, out of the enormous number reachable on the network, is the right one? Once that is answered, it has to answer a second question: which of the many services running on that computer is the one it wants? This article follows the answers to both questions. It starts with the protocols that carry the data and ends with the connection that the two programs finally share.

Keep these two questions in mind as you read. The first is answered by an IP address and, in many cases, a domain name. The second is answered by a port number. Together they identify a single connection.

Before looking at each piece in detail, trace the state a client program has to build up. Each step narrows down the target until one specific service on one specific computer is named.

The two halves do different jobs, and it helps to keep them apart. The Internet Protocol is the part concerned with identifying computers, which is why the addressing discussion in this article spends so much time on addresses. The Transmission Control Protocol is the part that provides reliable two-way communication between networked computers. Reliable two-way communication means data can travel in both directions between the two machines in a dependable manner.

The two parts are paired. When people say TCP/IP, they usually mean the whole arrangement rather than one protocol alone. Addressing and delivery between machines belong to the Internet Protocol side, while dependable communication in both directions belongs to the Transmission Control Protocol side.

Every computer on the Internet has an IP address that identifies it. You can think of the IP address as the machine's identifier on the network. A client program needs some way to specify which computer, among all those reachable on the network, it intends to talk to, and the IP address is the mechanism for that. Without a unique identifier per machine, the network would have no way to deliver data to one particular computer rather than another.

IPv4 addresses are 32-bit addresses written in dotted decimal form. Thirty-two bits are split into four groups of eight bits. Each group is written as a decimal number, and the four numbers are separated by dots. Since eight bits can represent values from 0 to 255, each part of a dotted decimal IPv4 address falls in that range.

IPv4 uses 32-bit addresses, while IPv6 uses 128-bit addresses. The difference is one of size: an IPv6 address has four times as many bits as an IPv4 address. The source pack establishes the bit sizes but does not describe how IPv6 addresses are written on the page, so this article does not give an IPv6 notation.

Numbers are efficient for machines but hard for people to remember. For that reason, many computers can be referred to by a domain name, a human-readable label such as math.hws.edu or www.whitehouse.gov. A domain name can be used instead of an IP address to say which computer you want.

A domain name and an IP address can both identify the same computer. The name is the human-friendly way to refer to the machine, and the numeric address is what the network uses to identify it. The source establishes that the name stands in for the numeric address. It does not describe the mechanism by which a name is turned into an address, so this article does not explain that process.

An IP address is enough to reach a machine, but a machine can host many services at once. A server does not simply listen for connections in general. It listens on a particular port. A port number is a 16-bit positive integer, and it selects which service on the machine the connection is aimed at.

This means the client has to know two things before it can connect: the Internet address or domain name of the computer where the server runs, and the port number on which that server is listening. Drop either one and the target is incomplete. The address says where to go, and the port says which door to knock on.

The diagram shows the two halves of the identity. The IP address picks out the computer, and the port number picks out the particular service on that computer. A connection is identified by both at once. Because the port number is part of that identity, one computer can carry many connections at the same time, each on its own port. That is what allows several programs to communicate simultaneously.

Standard Internet services have standard port numbers assigned to them. The standard port numbers are all less than 1024 and are reserved for particular services. The source's concrete case is the Web server, which generally listens on port 80.

If you write your own server program, the source advises using port numbers greater than 1024. That choice keeps your program clear of the range reserved for standard services.