A backup is a backup copy created by copying data files, directories or programs. Backups are used to restore the original data in case of data corruption or loss. The methods, storage media and frequency of backups depend on the timeliness and frequency of use.
Backups usually take place at the file level. A changed file is transferred in its entirety to the server and stored in mass storage. Depending on the backup strategy, the entire files can be saved, which brings significant speed advantages during recovery. In this case, we speak of volume-level backup or full backup. In full backup, complete 1:1 copies of the data source are created and stored on one or more backup media such as hard disks or, in the case of online backup, in cloud storage. Full backups provide the highest level of data security, but require a lot of storage resources.
If only the files that have changed since the last full backup are saved, this is known as a differential backup. In this case, only the files that have changed since the last full backup or that have been newly added are backed up. The newly backed up files are marked by the file system with special file attributes. A full backup is achieved by adding the last differential backup to the last full backup.
Incremental backup is different, where only the data that has changed from the previous backup is saved at regular intervals. Each backup is an image of the data changes between the backup source and the last incremental backup. Where the last backup can be a full backup or a differential backup.
Backups can also be implemented as cloud services: Backup-as-a-Service (BaaS). Backup-as-a-Service is comparable to Storage-as-a-Service( SaaS) where data is stored on the cloud storage of a BaaS provider.
Optimization before backup
If optimization of the files takes place before the backup, in which only the changed blocks of the files are transferred for backup, then this is data security at block level. This method is mainly used when transmission capacities are low. If the data backup takes place at a specific moment, it is referred to as a snapshot.
To circumvent the slow long-term archiving on tape drives during day-to-day operations, tapeless backups( TLB) can be performed as disk backup( D2D), disk to disk to tape ( D2D2T) or disk to disk to cloud( D2D2C). Deduplication technique can be used to reduce the amount of data.
The basic 3-2-1 backup concept
For reliable recovery of stored files, there is the 3-2-1 backup. This concept ensures that backup copies of stored data are always available when needed. The 3-2-1 rule refers to the number of backup copies and to the storage media used. According to this rule, there should be three copies of the data to be protected. In addition to the original data, two more copies should be created and stored on two different storage media and on an external storage medium, for example in a cloud. According to iT experts, this backup concept is considered best practice, where the greatest risks are eliminated.
With virtual machines, it is possible to mount backups. This term refers to the mounting of file systems. When mounting, the storage medium connected to the computer - the hard disk, USB stick, etc. - must be made known to it. - must be made known to it. To do this, the operating system reads some data and directories. During mounting, the virtual machine image is mapped on the backup server and can be put into operation directly, which significantly reduces the recovery time objective( RTO). If backup servers are used exclusively, then the applications run when mounted from the backup server, which imposes certain time constraints.
Backup systems are usually server-based, but serverless backups also exist. Typical backup storage media include tape drives, tape libraries, autoloaders, DVDs, HD-DVDs, Blu-Ray discs, disk arrays, jukeboxes and hard disks.