Networking

The basic advantages of networks are:-

·       the sharing of resources and information.

This can include the sharing of hardware such as printers, faxes, modems (or any other form of Internet connections) and storage and backup devices. The information could be documents or even ideas.

·       the provision of local facilities without loss of central control.

For large organizations central control of the computer facilities is most important. Users are monitored in their work, the software they use and accounting of the various network facilities such as printing can easily be recorded. Anti-virus software and only allowing the network manager to install new software prevents illegal software being installed. The ability to back-up all of the work from one server or station is also an advantage.

·       shared risk

If a terminal or printer fails then work should not be disrupted. Even cabling problems can be overcome by various "hub" architectures which can cope with one section of the network failing. Even network servers can be designed so that individual components fail (power supplies, hard drives and even processors) and yet the system remains intact. One method of achieving this goal is to use parallel systems with multiple processors and RAID hard drives. These are a made up of a number of smaller, cheap drives (when compared to traditional mini and mainframe hard drives.)

·       improved communications

Individual members of the organisation can have access to all other members. E-mail can be installed, messages can be broadcast, etc. The problem with E-mail is that it need not be read unlike mail that is delivered personally.

Software may be cheaper "overall" when shared between a number of users but extra network software needs to be acquired (For instance every time we add another computer to the school Windows 2003 network an extra client licence is required costing about £4). With application software a single user could buy MS Works for £25 but a 10-user pack across the network might only cost £150. However if only 8 users wish to use this package the 10-user pack would still need to be bought. Cabling and network cards, training of network staff and individual users all adds to the cost of implementing a network system. However, under most circumstances, the advantages overwhelm the disadvantages.

If you only have a small number of users and there is no need to share resources or data then stand-alone systems will probably provide a quicker response.

The following diagrams show the different common network architectures. Each of the circles in the diagrams could represent a terminal, a networked computer or a hub.

If these nodes are confined to a small area linked by fixed cable (thin Ethernet, optical fibre, etc) then we have a Local Area Network (LAN). If the nodes are separated by microwave, satellite or other non-permanent link over several (or several hundred) miles then we have a Wide Area Network (WAN).

Each of the systems has advantages and disadvantages in terms of costs and security. Multidrop and Ring networks are cheap but, if the cable is broken the whole system can collapse. Star and Mesh networks require more cable but if one segment fails the rest, in principle, should remain intact.

The use of HUBS (or, in more intelligent systems, SWITCHES) in which there is one input and several outputs transfers the architecture into a star form where each branch can split several more times - in effect a tree architecture.