Alex Pankratov :
Apr 08, 2019Hello Danielle,
That's a very broad subject. To each their own basically :)
There are two principal types of backups. There is a system backup and there's a data backup. The former is also interchangeably referred to as an "imaging" backup and the latter - as a "file" backup.
System backup treats the computer as a whole, so it grabs everything in sight and creates a snapshot of the whole thing. If the machine dies and you get a replacement, the idea is that you take a snapshot and restore the full setup back to its recorded state.
In comparison, data backup takes care of the data only. It's not used for backing up programs or Windows itself, just the documents. It is lighter and typically easier to manage.
Put differently, you'd use the system backup when you care about your system as a whole, and you'd use the file backup when the data is what matters the most. A typical example for the latter is safeguarding a photo collection.
These aren't mutually exclusive. It's perfectly normal to have both in place.
There is also hybrid backup software that is both system and data at the same time, which does full system snapshots and also makes individual files easily accessible if needs be.
The second aspect is how much of the past the backups should retain.
The simplest is to keep just a single replica of the source, a.k.a. a mirror backup. This is the default of Bvckup2.
Level up from here is to archive past copies (within the same backup job), in which case the program will retain older copies of all changed files and discard them after 2 weeks -
Another option is to run several mirror backups, running at increasingly larger intervals - hourly, daily, weekly.
Third option is to do a rolling backup by directing the same backup into one of several target folders and rotate these based on the current date, time, day of the week or the run number -
Third is how these past copies are stored. Once you start keeping older copies, there's an issue of efficiently storing or compressing duplicates.
This is a technical aspect, more of an implementation detail. Its importance is only in how much storage you will need and how long it will take to restore from a backup. Back-to-back backups are often stored as incremental deltas, meaning that retrieving certain version is a matter of "assembling" it from some full snapshot by overlaying relevant deltas, and this may take time.
With Bvckup2 everything is stored as is, in full. Given how affordable the storage space is nowadays, this works fine for the most of reasonably-sized backups.
How many _independent_ copies of the same backup is to keep.
This is typically governed by the 3-2-1 rule:
3) make at least three copies of your data
2) store two copies on different storage media
1) keep one of them offsite
That's about it.