What is a server?

A server is a computer like any other, but one which acts as a centralised, shared resource that other computers can access. Any computer can act as a server but usually when we talk of a server we are talking about a computer that has been specifically designed to act in this way.

Why use a server?

(The techie rationale)

To fully understand how a server fits into the overall IT infrastructure and why it is beneficial it's important to know how a network operates both with and without a server in place.

Peer-to-Peer Networks

The most common network infrastructure people will have come across is called a peer-to-peer network. In this arrangement each individual computer is connected to the others but no one computer is responsible for managing the network. If you share a broadband connection at home then this is most likely the method that is currently in use. Peer-to-peer networks are easy to set up, require little management and do allow basic sharing of resources between PCs, such as printers and scanners.

In a peer-to-peer network each computer has an equal role and an individual PC can share it's files or resources with others on the network. Whilst a peer-to-peer system is good for small networks it does have some significant disadvantages; the greatest of which is a lack of centralised management and control. Whilst this may seem like a minor point it becomes increasingly important as the network grows.

Imagine a peer-to-peer network of 30 computers, spread over an entire building, many of which have shared resources such as printers or shared folders. Keeping track of which resources are where and owned by whom would be a difficult task and making changes to the system would potentially involve accessing each of the 30 computers individually. Shared resources are also only available when the host PC is switched on which means critical resources could become unavailable if the user of that PC switches it off. Without some way to centrally manage shared resources the system quickly becomes chaotic and unmanageable.

The situation becomes even more complicated if you introduce security concerns into the equation. In a peer-to-peer network shared files are managed by the individual users of each computer which means implementing a network wide security policy is impossible. Critical business files are effectively distributed over the entire network and as such there is no way to ensure data is backed up and secure.

The last significant drawback to peer-to-peer networks is that these networks simply don't allow for many of the features currently demanded by modern businesses. Company e-mail systems, remote working solutions and many line of business applications all require networks with a central 'supervisor' to manage those resources.  

Server-Client Networks

To address the problems most modern business networks use what is called a server-client relationship. In this system a server is introduced which effectively manages any shared resources and in many cases the actual network infrastructure itself. As the diagram shows, the physical structure of the network changes very little with the introduction of a server; what has changed is the functional structure of the network.

In this arrangement the PCs are now clients of the server rather than equal partners in the network topology. Resources such as printers and shared files are no longer controlled by the individual PCs but rather by the server. This facilitates centralised management, backup and security and also enables resources to be allocated far more efficiently. In our example network, making changes is now easy because all the shared resources are controlled from one location.

Why use a server?

(The business rationale)

The bottom line is a server greatly reduces the total cost of ownership, improves security and vastly simplifies the management of the network. As a business owner this means several things:

• Less Downtime - Simplicity of management means less downtime and more productivity. A well implemented server network will mean less lost days and less IT headaches when compared to a workgroup of the same size.

• More Secure IT - Servers massively improve security and reduce the risk of accidental or malicious loss of data by implementing centralised security policies.

• Less Lost Data - Servers enable centralised storage of data and facilitate more streamlined backup processed. This means less lost data, better disaster recovery and tighter compliance with data protection regulations.

• More Productivity - Features such as shared mailboxes, shared calendars, global contact lists, intranets and public file stores make collaborative business a reality. This increases productivity, creates closer more efficient teams and streamlines company communications.

• Less Expense - Whilst a mixed bag of ad hoc solutions might offer similar features to a server the total cost of ownership is almost always higher when you factor in maintenance costs and downtime.

• A more scalable IT infrastructure - Workgroup networks scale very badly when 10 or more client computers are connected on the network. This results in poor network performance, glitches and downtime. Server based infrastructures can comfortably scale to hundreds and thousands of client computers so your IT will grow with your business.