Back

Backing up versus archiving data

Backing up data
Have you ever lost work you have done on the computer? Do you backup your work onto a pen drive every time you do some work on the computer? If you don't, you should! Given the amount of coursework you probably have to do, it is a wise approach keeping up-to-date backup copies of work that goes towards your final course grades! Your teacher would have told you to do this many times and you will get little sympathy from anyone if you lose work! Companies must also back up their work. If you lose your work, you can start again - not much fun but possible. If a company loses files they could go out of business. People could lose livelihoods. Data is valuable to an organisation. It takes companies years to build up a customer base. Getting data into a computer takes time. Businesses today are run on computers. Examples of records that a company keeps include:

    • Employees' details, so they can be paid.
    • Details of who owes the company money, so the company can collect payments, pay their bills and make a profit.
    • Details of who needs to be paid, so that they don't stop supplying you with raw materials and services because they haven't been paid.
    • Designs of products made with CAD.
    • Correspondence with suppliers and customers.
    • Legal information.
    • Web page designs.

What might cause a company to lose its valuable data?
This could happen in a number of ways.

    • Hardware failure.
    • Corruption of files.
    • Attack by a virus.
    • Attack by a hacker.
    • Theft of data.
    • Accidental misplacing of data.
    • Theft of the equipment that data is on.
    • Sabotage by an employee.
    • An honest blunder by an employee.
    • Espionage by a rival company.
    • A natural disaster.

A backup strategy
An organisation needs to implement a backup strategy so that if data is lost, the activities of the organisation aren't compromised. A sensible organisation will have a backup policy written down. This will deal with the following things:

    • The hardware/software that will be used to backup a system.
    • The person responsible for backups.
    • The frequency that backups will be done.
    • When they will be done.
    • Where backup copies will be kept.
    • How backup copies will be labelled.
    • How the company will know the backup system is working as it should.

A backup system needs to use media that is big enough to hold all the data to be backed-up! Magnetic tape is often used for backing-up networks because gigabytes of data can be stored on them. Servers often have a special piece of hardware in them that holds a number of magnetic tapes. The software that comes with it can be configured to automatically backup data in the middle of the night and eject the tape ready for the network manager to put in the fire safe first thing in the morning. Backing-up networks to cloud storage is also increasingly common. Files are stored off-site, with many different versions being stored going back in time. The cloud storage company takes responsibility for the safe storage of data, ensuring it is protected from hackers or viruses, for example. Of course, you need a reliable Internet connection for this system to work. USB pen drives, CD R/W or cloud storage are all useful backup storage media to use for personal, day-to-day files.

Backups need to be done, or at least managed, by a named person so that everyone is clear whose job it is to do what and whose job it is to check that the system is working as it should. A backup file quickly becomes out-of-date because the actual system is constantly changing. Therefore, backups need to be taken frequently enough to ensure that only a very limited amount of data is lost. Because a 'rolling program' of backup copies will most likely be made, it is important that everyone is clear how these will be labelled and where they will be kept. Will a copy be kept in a fire safe or off-site, for example? If off-site, how often will it be changed and who will do it? It is very easy for this kind of system not to be done properly because backup copies are rarely needed. Systems generally run smoothly. The problem comes when there is a problem!

Archiving data
Although often confused by students because they seem to involve the same sort of tasks, they are completely different terms used to describe completely different things.

Computer systems invariably take in data. Consider a school. Each year, new pupils join a school and some pupils leave. If all a school ever did were to input the details of pupils into the system but never remove details, the system's resources such as hard disk space would soon diminish. Therefore, the school should remove pupils who leave. Consider a company. They may keep accounts, orders, invoices and the like on the organisation's network. Once a company has received the goods it has ordered, the order, invoice and any other associated paperwork may be redundant. If you keep all of the computer files a company makes on the system but never remove them, it won't be too long before the system's resources are under pressure. There is a problem, though. You cannot just delete pupils' details from school once they have left. Neither can you destroy invoices. A school is under a legal obligation to keep the records of pupils for many years after they have left. What would happen if a pupil needed a reference? Any company must keep financial records for many years after they have become redundant - the tax office might want to check details of tax returns years after a tax year has finished. Consider what happens after a serious accident involving a plane crash a space shuttle. An accident investigator might want to trace components used in the aircraft or shuttle. They can only do this if records are kept of each component for many years after they were actually made. You want to get rid of data because it is using up hard disk space and slowing down the retrieval of active documents - it takes longer to search a big file than a small one, for example. But for various reasons, you can’t! What should the organisation do? The answer is to ‘archive’ the data.

    • There should be a written archiving policy in an organisation stating who is responsible for ensuring data is archived, intervals between archiving and so on.
    • Archiving should take place at appropriate time intervals, e.g. every 6 months. Compare this to backing-up files.
    • Redundant data could be copied onto a magnetic tape or other mass storage device because you can store more bytes per unit cost compared to other storage devices.
    • The data may also be compressed so you can fit more of it onto the tape.
    • The archived data is meant to be kept for a long time and doesn't generally need to be easily accessible.
    • The original data can be deleted from the system once it is archived.
    • The magnetic tape should be labelled up and put in a fire safe or somewhere else secure.

In summary, archived data is generally data that is unlikely to be needed again but you can't get rid of it for legal or other reasons. You want to remove old data from a system because it ties up system resources. Should you need to get back data that has been archived, it will be possible but may take a little bit of effort. You will have to get the correct archived tape, un-compress the data if it has been compressed and then find the data on a serial file, which can take time.

Back