On the opposite end of the spectrum to fault tolerance, Disaster recovery, as its name suggests, is the ability to get a system functional after a total system failure (clearly a costly and time consuming concern for any modern business) in the least amount of time. Strictly speaking, if enough fault tolerance methods are in place, you shouldn’t need disaster recovery, but of course this cannot be taken for granted.

Disaster recovery will mainly involve comprehensive backups of your data at short regular intervals. A company’s data is often their most valuable asset, and any IT professional will state that the most common hardware failure is the hard drive.

These backups should be in a variety of different media to ensure your data is safe, although obviously this is not always feasible. Depending on the amount of data and its importance you should set a regular timetable in which your hard drive backups are transferred to optical media, solid state drives or magnetic tape.

Replacement hardware should also be kept in reserve to reduce waiting times for new components. The priority for disaster recovery is to get your system up and running again as quickly as possible so any competent systems administrator will be ready to replace hardware and software at short notice.

The Domain Name System (or DNS) converts the Internet domain and host names into IP addresses. DNS by design translates the address a user enters into a web browser to the IP addresses of the servers where the site is situated. An address using the alphabet is much easier for a user to remember than a numerical IP address, so DNS is a vital system that was pivotal in bringing the Internet to the mass public.

DNS implements a database to retain this name and address information for every public host on the Internet. DNS works on the assumption that IP addresses are static rather than dynamic. The DNS database is kept on a system of database servers. When Web browsers issues a request involving Internet host names, software known as a DNS resolver (usually built into the network OS) is utilised. The software first contacts a DNS server to determine the server’s IP address. If the DNS server does not contain the required IP address, the request is forwarded to another DNS server. This process potentially repeats several times until the IP is given, which is then sent to the DNS resolver. The resolver translates the request over Internet Protocol and the web browser is finally able to view the site.