Backup

Data backup (english backup [ bækʌp ] ) refers to the copying of data with the intent of being able to copy them in the event of data loss.

The saved on the storage media data as a backup, Eng. Backup designated. The restoration of the original data from a backup copy is referred to as data recovery, data restoration or (english) Restore.

• 5.1 Example

• 8.1 Hot Backup
• 8.2 Cold Backup

• 9.1 criteria 9.1.1 The type of data
• 9.1.2 The value of the data
• 9.1.3 The change in frequency of data
• 9.1.4 Legal requirements
• 9.1.5 location
• 9.1.6 time of data backup
• 9.1.7 requirements

Implementation

The storage of data backups should be done locally from the computer system and in a safe environment.

• For private individuals to provide external hard drives with FireWire, eSATA or USB port. These can be easily connected to the system to be secured and separated again from this and allow at least one remote storage. Also, network-based hard drives are easy to connect and remove and therefore efficient backups possible.
• For smaller companies, for example, bank lockers suitable for disk storage. However it can not be accessed at any time it is usually because the access to the data volumes is only possible during the opening times of the bank. An alternative to this is the online data backup is: the backup takes place outside the home, usually in a data center, and it can always be accessed. In this case, however, is to make sure that the data transfer is done in a safe manner; the external service provider should not be able to read the contents.
• For larger companies specifically secured safes or premises may be worth (called cells) to fire- safe storage of tape library. The backed up data can be distributed to multiple sites or data centers.

Legislation

The obligation to provide backup in enterprises arises inter alia, the legal provisions on a proper, traceable, auditable accounting (HGB). From the short -term storage ( limited to single tag to three or even six months), the longer -term data archiving, which is subject to different laws differ. The principles for archiving and verifiability of digital assets are in Germany since January 2002 for companies operating in the Principles of Data Access and Verifiability of Digital Documents (GDPdU ), published by the Federal Ministry of Finance, together.

Documentation

In the backup, it is very important to keep good documentation, since the success and speed of data protection and recovery may depend on it.

The documentation should include:

• Expiration of backup
• Structure of the archiving
• To be taken (immediate ) measures
• Competencies ( the employees and service providers )
• Priorities for particularly time-critical data and systems

For better clarity, the documentation for backup and recovery is to be determined separately in a backup and recovery plan.

Backup Types

Depending on the change in intensity of data to back up certain types of backups can be used in the actual backup process. Single backups can full backup, differential and incremental backup can be distinguished. Differential and incremental backup requires at least one full backup was performed. In the normal backup specific files and / or directories are selected (folder) whose content is to be secured. There is also the possibility to save certain file formats only. In addition, even entire hard disks or partitions can save it as an image. In all cases it is possible to also restore only part of a complete backup set.

It is divided into:

Komplett-/Vollsicherung

The complete or full backup is referred to in programs as " Normal Backup". Here, the respective data to be saved ( an entire drive, a partition, certain directories and / or specific files, specific file formats ) are completely transferred to the backup medium and marked as backed up.

Dump backup

In the memory dump backup ( english image backup), the complete disk secured (usually the hard drive, but also a USB mass storage, optical media or in some programs also device on the network ) or just a partition through a 1 - to-1 image be. For example, not only the user data, but the entire file system, including the operating system and user settings saved. The advantage of this assurance is that in case of total failure of the computer memory dump saved back to disk and thereby the status of each disk can be fully restored to the backup time. In such a recovery is either the entire file system in its original structure restored ( in this case, is not a file system driver is required, but only a device driver for the disk access ), or a special driver reads regularly the file system and extracts only the desired files and directories from the backup to integrate them as normal directories and files in the current file system or the current to overwrite the older saved (see " Incremental Backup "). For several years, programs on the market that can create incremental backups those also.

Differential Backup

In the so-called differential backup all the data that have changed since the last full backup or added, saved. It is thus placed repeatedly on the last full backup, with respect to a new full backup space and time can be saved.

Incremental Backup

An incremental backup only the files are stored, since the last incremental backup, or ( at the first incremental backup ) have changed since the last full backup or added. It is therefore always placed on the last incremental backup. This method has the disadvantage that when restoring the data usually have to be culled from several fuses.

Detection of changed files

To distinguish them from changing files and previously saved, unchanged files, some file systems use special file attributes, which are set automatically by the system when a change and are deleted when a full or incremental backup again from the backup program (such as the archive for FAT and NTFS). If such attributes are not available, the backup software on the files and their timestamps must keep records (eg tar snaphots under Linux).

Towers of Hanoi

In order to achieve a good compromise in the number of data retained backups and made ​​available to end hardware, it also uses the Fuse " Towers of Hanoi ". This backup strategy is based on the eponymous puzzle game. Each used backup medium corresponds to a slice of the towers and with every movement of the disk backup to the appropriate media is played. Thus, the first medium is used every other day (1, 3, 5, 7, 9, ... ), the second one in four (2, 6, 10, ... ) and the third one in eight ( 4, 12, 20, ...).

With n media can thus get by 2 n- 1 days, until the last media is overwritten. Thus one has in three media nor backups of four days ago, on the fifth day the backup C is overwritten. At four media has eight days until the ninth day medium D is overwritten and five media has 16 days at 17, the medium to E is overwritten, etc. files respectively, pending upon the amount of media, before 1, 2 4, 8, 16, ..., 2 n-1 days to recover. Mathematically, the medium to be used is determined by the number of zeros on the right side of the binary representation of days since the beginning of the backup.

The following tables show which media are the days used in different number of media. It should be noted that in this method, the first backup will be overwritten after just two days. However, this can be avoided by starting (marked in red in the Tables ) at the end of the cycle.

Towers of Hanoi for three media

Towers of Hanoi for four media

Towers of Hanoi for five media

Special case of private users

For home users the kind of sensible data backup depends strongly on the means available to available hardware, the existing knowledge as well as on the personal attitude to the data being backed up and the backup. With sufficient commitment can already create in simple backups and remove the safety on industrial level.

In the software market, both commercial and free programs are available. Among the best known commercial offerings include programs Acronis True image of the company, the company StorageCraft ShadowProtect, DriveImage XML from runtime software, Carbon Copy Cloner for Mac OS X from Bombich software. In the freeware area Cobian or Areca may be cited as examples, but even simple tools like robocopy or rsync from Microsoft and its resulting data backup solution, rsnapshot under UNIX. Starting with Mac OS X 10.5 (Leopard ) is integrated into the operating system with Time Machine, an automated backup solution for backups on external hard drives (USB / FireWire or network drives ).

Example

It makes sense to backup on a separate hard drive. An external hard drive can be kept in a safe place after backup from the computer. In an internal hard drive has to be at least that any viruses and malicious programs do not have write access to the backup media during regular operation. Hard drives with huge storage capacity are becoming cheaper. Even on a USB memory stick or DVD / DVD -RW backups are impractical. Burner for laptops and desktop PCs include long been the basic amenities and the blank media are favorable. The easiest way to create a pretty good backup without software and with little background knowledge, is investing at least two fuses at the regular interval to physically independent media. So the grandfather - father-son principle can be simulated. With three or more media can expand this principle to the effect of being able to make kleinschrittige undo changes or further vorzuhalten past versions. In other media can be the speed and capacity increase.

If the data on the original hard drive sorted according to the backup current or most important data can be done in shorter time intervals ( eg, daily ) than those of other stocks.

History

In the 1990s, Iomega tried the Zip disks with - for that time - comparatively high capacity of 100, later to be positioned up to 750 megabytes in data protection solutions. Magnetic tapes have extremely low in the private sector and dissemination are the hard drives to speed and especially in the cost per space now inferior. In the energy consumption as well as in durability, they are superior but what they can still be made ​​in the corporate use. Hard drives now offer large capacity and relatively stable prices devices an attractive alternative to removable media. Also flash memory have reached workable capacities and are therefore suitable as backup media.

Media types of data backup

In 2005, most backups are made ​​(eg, digital linear tape, Linear Tape Open), hard disk or optical storage such as CD- R, DVD, DVD-RAM and comparable formats of disk-based production systems to tape a large capacity. With the increase in cheaper broadband Internet connections gain network and online data backups on external servers more important.

In the private sector, other backup media are used (see special case of home users ).

Real -time applications

Databases must be backed up in a consistent state (data consistency, see also Database Archiving). This can be achieved by the database is shut down, then performed a data export and then the database is started up again. This method is called in the jargon also cold backup, since in this case - in contrast to the hot backup - the database separately from the production network ( ie such as the Internet or intranet) and thereby the operation is interrupted.

Hot Backup

A hot backup is a backup of a system (eg a database) that is created during operation of the system. This allows a backup to be kept as current as possible - ideally, it is on the same level as the live system. Advantage of this method is the Provision of a current " replacement data set ," which is immediately ready for use in case of a system crash. In most cases, the system must support this backup method, otherwise a fuse due to the active use may have inconsistencies. This procedure is also called Online Backup.

Cold backup

A cold backup is a backup of a real-time system that is created while the system is not active. This ensures that the data is saved in a consistent state. The disadvantage of this method is that the system is not available for the period of time of the backup. For high-availability services it is therefore inappropriate. To create copies of protective environments, for example, only during the day must be available, it makes sense to contrast. This procedure is also called offline backup.

A common method for Oracle databases is to put the database at the start of the backup in the data storage mode and then back to production mode.

Different manufacturers of backup software and other manufacturers offer online integrations ( Integration Agent ) and additional products such as the Open File Manager of St. Bernard.

Data Backup Strategy

A backup strategy can be used anywhere where there are unique data of a certain value, whether in the private user area, in projects or in the corporate sector. In the latter case this can exist as a binding stipulation in the form of a directive.

In it can be set:

• What has to be done the backup.
• Who is responsible for data backup.
• Be performed when data backups.
• What data should be backed up.
• The storage medium is to be used.
• Where the backup is kept secure.
• How to back up the data before data theft (for example, by encryption).
• How long backups are kept.
• When and how backups are tested for recoverability.

It should also be determined when and whether (a) a full backup (eg on weekends) and / or ( b ) an incremental or differential backup (weekdays eg at midnight ) performed.

Criteria

The optimal backup strategy depends on many factors and therefore be re-established in each individual case. Important factors which must be considered are:

The type of data

The value of the data

Here are three aspects to be distinguished: First, what loss is caused when the data is irretrievably destroyed? For example, in an enterprise data is backed up on a daily basis in the night, all observations must be repeated at a loss of data just before closing time. From the work of the employees concerned there is a clue for the loss. Especially with the irreplaceable data, however, is often the ideal value to be considered.

Secondly, what loss is caused by the time required for full recovery and, where applicable, can not be used? For example, if the installation of a PC takes one day, the loss may exceed the value of the software installed far. Here you would select a backup method that makes it possible to fully reconstruct the installed state very quickly (memory dump ).

Third, the costs arising from the duty to provide information that is may in accordance with the Federal Data Protection Act or the laws of other states? When data loss of certain types of personal data, the concerned, the supervisory authorities or the public should be informed of the data breach.

The change in frequency of data

This factor has a significant impact on the use and design of the generation principle. Data with low frequency of changes, such as operating system and installed software may not necessarily be backed up regularly. It can also be sufficient to secure that only areas before or after surgery.

The faster data can be changed, the less you will choose the duration of the backup cycles corresponding to the generation principle. Note this is also the time to expiration. While statutory retention periods are for a lot of data in business ( for example, invoice data ), such as the current contents of Web pages may be discarded after a short time when they are no longer needed.

Legal requirements

The data backup strategy must be able to guarantee any regulatory requirements (eg audit compliance).

Note the generally accepted principles of computerized accounting systems, specifically paragraph 5.1 and 5.2.

Location

Thus, since there are very different kinds of data with different requirements on the backup strategy, it is useful this data in advance to different locations (hard drives, partitions) to separate. Then the optimal strategy can be selected for each location. Additionally, there are accident- only data storage. With online backup, the data in most cases are stored in a data center.

Time the backup

When choosing a suitable concept in particular the time required for data backup is important from a business perspective. The total cost consists of the recurring security costs and in the event of data loss incurred costs of restoration. The relation in which each other, these two variables is dependent on the selection of a specific backup procedure. A low assurance effort is particularly sought if large amounts of data must be locked during the backup process, which can be for decades but often avoided in many systems. There are for this purpose software may back up the data of a system in operation.

Requirements

Depending on the medium and type of data backup, the criteria will vary widely. However, are mostly mentioned the following points:

• The procedure in the case of emergency should be known to several employees. A checklist for this case is very useful because in an emergency, often no one has time or nerve to think what to do next.
• Where possible, the data before backup should not be compressed. Redundancy can be useful in recovering data.
• It is at least willing to hold a drive that can read the media used.
• The economic benefits of backups (cost to restore the data without backup ) should be in reasonable proportion to the operated for backup effort.
• The only reliable evidence of a successful backup is to prove that the backed up data can be restored completely and within a reasonable time. For this reason, restore testing should be done at regular intervals.