In today’s world, data is a valuable asset and for this reason requires a particularly high level of protection. Data can be lost in various ways, e.g. through unintentional deletion, technical defects, attacks by hackers or elementary damage (fire or flood). Therefore, the use of data backup attempts to prevent irretrievable loss. The task of data backup is not only to counteract data loss, but also to ensure fast recovery.
For this reason, the following golden rules of data backup should be followed. These rules are primarily applied within companies, but can also be implemented to some extent by private individuals.
Develop data backup concept
Before the actual data backup can be started, a data backup concept must first be developed. This defines all the important prerequisites for subsequent data backup, both technically and organizationally.
It is determined in this phase which criteria must be met so that a backup of all relevant data is possible without complications.
Points that are important for the later data backup would be, for example, the determination of the data backup system used, the hard disk contents to be backed up or other general conditions.
Determine responsibilities / accountability
It must be clearly defined which persons within a company are responsible for data backup. Failure to clearly define responsibility for data protection is a common mistake that many organizations make.
It is recommended to assign at least two people the responsibility for the backup, so that one person is always available in case the other is on vacation, for example.
These employees must be specially trained for the field of data backup so that it is known how, what and where data is backed up.
Automation
Data backup should be automated – to the extent possible for individual companies. Unfortunately, many medium-sized companies do not have automated data backup processes, which means that, for example, the change of data carriers to which backups are made is forgotten or that the actual backup process cannot even be started manually.
For this reason, the manual steps of data backup should be automated as much as possible. In this way, it can be ensured that the backup process is carried out regularly at a specified time.
Particularly helpful for this task is the use of a conveyor robot. This automatically inserts the appropriate tape at a previously defined time and can thus be used for different data backup methods.
The use of robots or other automated solutions increases the effectiveness of data backup and ensures regular execution of the backup process.
Define datasets to be backed up
Before the actual data backup can begin, all of the company’s data assets must be categorized. This is especially important to ensure that only data that is important to the company is backed up and to avoid resource bottlenecks on the backup media.
Important data and files whose content changes frequently should be backed up daily. All other records should be captured at least once a week as part of a full backup.
Additionally, care should be taken to ensure that all data that is to be backed up is stored in a central location so that the backup can be performed. Individual content that resides locally on individual PCs is cumbersome to back up and increases susceptibility to errors.
Set frequency of data backup
When planning how often to back up data, it is important to consider that data that changes at short intervals is backed up more often than data sets that are not used for long periods of time.
The intervals between backups should be fixed periods of time that are repeated periodically. For example, there could be a daily backup of frequently used directories and files, and a weekly backup in which the entire data set is backed up.
Additionally, it is advisable to perform other unscheduled backups, especially before major changes or events occur.
Set backup type
In the application area of data backup, different backup types exist to ensure that all data is backed up at recurring intervals.
Backup types
Differential backup:
Differential backup backs up all files that have been changed or added since the last backup of the complete system. This type of backup has the advantage that the required storage space per backup is relatively small compared to other methods and the backup can be performed in a shorter time. The last complete backup is always the deciding factor as to which files will be saved during the next backup.
Incremental backup:
Incremental backup, similar to differential backup, saves all files that have been modified or newly created. As a basis to determine changed files, the first incremental backup compares the records with the last full backup. In subsequent backups, the last incremental backup is used as a reference and only files that have been manipulated since then are backed up. In this way, a kind of fuse chain is created in which each fuse rests on the last. Therefore, with this type of backup, it is especially important that all backup steps are successful so that a complete restoration of files is possible.
Define backup strategy
Different strategies are used to manage the versions of the disks used in data backup. These strategies then determine at which point in time the individual data carriers are overwritten again.
Backup strategies
First in, first out:
The backup strategy FIFO represents a very simple form of data usage on the storage medium. As soon as the space on the storage medium is almost exhausted, the oldest complete backup is overwritten.
Grandfather-Father-Son:
This backup strategy relies on the generation principle, which is very popular and used by many companies. The “grandfather”, the “father” and the “son” each represent a point in time at which the data is saved.
The “Son” backup is performed every day, the “Father” backup is performed at the end of the week and the “Grandfather” backup is performed at the end of each month. For example, if backing up to LTO tapes, then you need between 4-6 tapes for the “son”, four tapes for the “father” and any number of “grandfather” tapes for the end of the month.
Encryption
When performing data backups, confidential data should always be stored only in encrypted form on the respective storage media.
Encryption can usually be performed directly during the backup process using the data backup software in use.
This protects the secured data not only in the event of loss, e.g. through theft, but also against unauthorized access by employees.
Verification of the data storage
The storage media used for data backup should be checked regularly for defects and error-free functionality so that only storage media that are in proper condition and function reliably are used.
For this reason, different or multiple media should be used to back up the data so that if one media is defective, the other backups can be performed correctly.
The best data backup concept is useless if the storage media used are not functional to the extent required.
Storage of the data media
It is best to store the backup copies in a different building than the original data media; this could be in a safe deposit box at a bank, for example.
If this is not possible, then the backup copy should at least be stored in another fire compartment.
Test reconstruction
Regular tests must be performed to verify that the media created during backups are ready for use and that data can be reconstructed without problems.
These checks are important so that problems with the data backup are not detected only in the event of an emergency. In addition, the tests can be used to verify that all desired IT systems are included in the backup.
Especially the IT infrastructure within many companies changes over time, old systems are replaced and new ones are set up. Data protection must be adapted to these changes, and regular checks make misconfigurations more likely to be noticed.
Related links
BSI: Data backup concept
BSI for Citizens: Data backup
Founder’s Encyclopedia: Data backup also a lifesaver for small businesses
Wikipedia: Data backup