By David Fox
A Backup writes data onto ‘containers’ which can be tapes or large files on disk/cloud. Over time, many different versions of files will be saved into these containers. The containers are filled up and are ‘recycled’ over time, removing older backed up data to make room for new. Restoring from the backup requires using the backup software that wrote into the containers.
Alternatively, Cloning/Replication is simply a copy of the source files and folders onto another disk. The cloned data is accessible, in the same way as the original data, it’s just a copy created for backup purposes. Once the initial clone task is completed, it can be updated by identifying changes to the source files and overwriting at the destination, keeping the clone up to date.
In this article, we’ll focus on backups, and examine three different techniques—full, incremental, and differential—backup software uses to save data into the container files.
Types of Backups: Incremental, Differential & Full
The first of the three types of backups we’ll discuss is full. A full backup saves all source data into the storage containers. Full backups provide the foundation upon which incremental and differential backups sit. Full backups can be time-consuming but are necessary. It’s usually necessary to repeat full backups at intervals to provide a fresh starting point for the other types of backups we’ll discuss below.
Incremental requires that a full backup has already been performed. Incremental backups save all changed files since the last backup. This type of backup is usually run daily to add the day’s changes to the backup. Restoring requires reading from the data stored by the full backup plus each incremental up to the point the restored data is required from.
In the graphic below, we see a full backup was taken on Sunday, and each day an incremental backup is taken, adding the day’s changes to the backup. Think of the incremental backup adding layers of data on top of the full backup.
The last of the types of backups that we’ll discuss is differential. A differential backup is an alternative to incremental backups. Differential backups save all changed files since the last full backup. This type of backup usually runs daily. Each differential backup saves all changes since the full, so they get larger over time.
The graphic below shows each day’s differential backup. Each one is larger than the last because it includes further days worth of changes.
Since it rarely makes sense to use incremental and differential backups together, one must decide which makes the most sense in a given scenario. Incremental backups are an efficient use of storage resources and time because data is only saved once. A file created on a Monday will only be saved by the incremental backup that runs on a Monday evening and need not be saved again until the next full backup saves everything again.
In contrast with a differential backup where the same file from Monday is saved in all differential backups from Monday onwards. Not such efficient use of time and resources – especially in environments where there is a lot of data to backup. In addition, a move to using cloud storage works against differential backups, where the additional storage overhead translates into increased costs.
The benefit of differential backups is felt most when restoring. Recovering all files that existed on Thursday only requires reading the full backup and the differential backup from Thursday. Contrast with restoring where incremental backups are taken – the same restore would require restoring some data from both the full and the Mon, Tues, Weds and Thu incremental backups.
Protect Your Data with Archiware
Archiware P5 Backup uses full and incremental backups to achieve reliable and accessible backup. Commonly used by media companies, where data-sets are often massive, the benefits of differential backups are outweighed by the efficiency and speed of incremental backups. Daily re-saving visual/video/audio data is not an option in media environments.
Add to this Archiware P5’s sophisticated indexing of saved data that efficiently tracks the layered incremental backups and optimises accessing of data when restoring, and we have a winning combination.
If cloning/replication is required Archiware P5 Synchronize can be used to copy a set of source files and folders onto another disk. The target storage can be located locally or at another site or data-center. P5 Synchronize includes a scheduler to regularly repeat updating of the target data. Previous versions of files or entire snapshots of the data can also be maintained.