The Internet has become, arguably, the most important and pervasive example of a network on the planet. The Internet connects people from across a street, and from around the globe at nearly the speed of light. The sections in this chapter will discuss some fundamentals about the internet, and some more advanced chapters on the subject will be discussed later in the book.
The exact relationship between a client and a server in the traditional Client-Server relationship can be complicated, and in today's world, the distinctions are more hazy still. This page will attempt to define what a client and a server are, how they are related, and how they communicate.
- A client is a communications entity that requests information
- A server is a communications entity that supplies information, or responds to a request.
In the page on Time-Division Multiplexing (TDM), we talked about how we can use a method of breaking information up into chunks, and prefixing that data with address information to send multiple streams of data over a single channel. These chunks of data with their headers are called packets, and are the basis for communication in the Internet.
In local area networks (LAN), packets are sent over baseband channels for very short distances. These networks use Statistical TDM (packets) to control access to the channel, and to control the process of moving information from its source to its address. This process of gettings things where they need to go is called routing. Devices that route packets are called (surprise!) routers.
Over larger networks such as Wide Area Networks (WAN), high-capacity broadband channels such as fiber optic cables connect different local LANs. Over a WAN network, packets are then Frequency Division Multiplexed (FDM) to flow simultaneously over these broad channels. At the other end, these packets are moved back down to a baseband system, and routed using TDM techniques again.
When talking about the different components in a computer network, a lot of different words fly around: Routers, Hubs, Switches, and Gateways, are some example. This page will talk about some of the different pieces of hardware that make the Internet possible. This page will only offer brief explanations of each component, opting instead to save complex discussions on the workings of each for later chapters (or even later books!).
An Ethernet hub, normally just called a hub, is a networking device used to connect multiple Ethernet segments in order to create a primitive LAN. They are primarily connected using unshielded twisted pairs/shielded twisted pairs (UTP/STP) or Fiber Optic wires, and require no special administration to function. Hubs operate on the physical layer (Layer 1) of the OSI model, and indiscriminately forward frames to every other user in the domain.
Hubs perform a variety of tasks, including:
- Acting as a multi-port repeater,
- Forwarding frames to all users,
- Allowing for connectivity to networks using 10base2 or 10base5 media,
- Partitioning ports when excessive collisions occur, which isolates the affected segments,
- And broadcasting a "jam" signal when collisions take place on Ethernet media.
Hubs also extend but do not control collision domains, absorbing bandwidth and allowing excessive collisions to occur and hinder performance, when switches or bridges can effectively break up a network.
Despite the increase of switches as the connection medium for workgroups, hubs can still operate in a number of situations:
- Small computer clusters,
- Careless users/saboteurs cannot interfere with a hub's function,
- Connecting outdated networks to newer methods of data transmission,
- And when budget is a priority over functionality.