Backup theory

Not all backups are made equal; there are different types of backups depending on what the approach is and what the aim is.

The type of backup can alter the type of transfer and how it’s stored, so it’s worth having a quick blast through the approaches.

We’ve been largely concerned with active backups in this article, so it makes sense to cover those first. As the name suggests, these are duplicate files of actively used system or data files. These contrast with archive backups, which we’ll cover shortly. An active backup attempts to back up all files, and then maintain an up-to- date version or versions of these according to a schedule selected by the user.

Archive backups are only concerned with data that’s no longer actively used, but copies are required to be retained. They’re essential for businesses, but also for individuals who want to safely retain financial or legal records, message logs and so on. They require a different approach to that of active backups in terms of reliable long-term storage.

Synchronised files are a modern spin on the active backup, and enable a file to be accessed from any location, or remain the same in multiple locations. Dropbox is a classic example, but modern ‘office’ services, such as Google Drive, provide something similar, with multi-user access and versioning.

Synchronised services typically require an internet connection to run the backup and be accessed, unless a local service is being run, such as OwnCloud or Nextcloud. An active backup is best run over a local network or wired drive, though for more limited purposes, a cloud service would be adequate. You would want to consider redundancy on the backup medium, such as RAID. Archives would ideally be stored offsite — a cloud service would be best — and if the data is irreplaceable, a bitrot-aware filesystem, such as ZFS or Btfs, would be advisable.