One thing that often crops up when discussing backups in general is full backup – a backup ‘level’ that creates a copy of all of your files or disk sectors and creates an image file. While this is one of the most straightforward forms of backup form, there are also some downsides, from limited flexibility to the potentially enormous amount of time required to perform a backup job and associated potentially large storage space requirements.
The Full backup level is not always the best one, but it’s still often used as a part of a company’s backup system in general. The most common time period between full backups in most organizations is once per week, with additional full backups carried out on a case-by-case basis (for example, if there’s been a software or system upgrade).However, even a week’s worth of data that a company could lose in the event of a disaster would mean colossal financial loss for many organizations. That’s where other backup types come in – ones that are more efficient and more flexible.
The majority of data on an average computer rarely, if ever, changes. This applies to apps, user data and the OS itself. Most of the time there’s only a small percentage of data that changes each day (or each week). With that in mind, it makes sense to only back up just the specific data that changed since the last backup. This approach forms the basis for a lot of the more complicated backup levels, but we’ll be discussing only two of the most popular ones: differential and incremental.
Differential backup was the first step towards backup level improvement. This backup level is all about backing up any data that has changed since the last full backup was performed.
An example of this strategy could be the following: if your last Full backup was on Monday, on Tuesday your differential backup software would back up all data that had changed since Monday. On Wednesday your system would also back up all of the data that had changed since Monday, and so on.
While Differential backup is likely to be much faster than your regular Full backup, it is probably still going to become quite storage-heavy, because the total amount of data being backed up accumulates with each additional differential backup - until the next Full backup is done. The optimal time period between Differential backups may be (for example) once a day at most companies, due to the amount of storage it takes for each new backup that your Differential backup software manages and creates.
Incremental backup differs from a Full in that it backs up the data that changed since the last backup happened (it does not matter if it was a Full or another Incremental). The other name of this backup strategy is “Differential Incremental Backup”, and the Differential backup is sometimes referred to as “Cumulative Incremental Backup”.
To go with our previous example one more time: if there was a full backup on Monday, on Tuesday your incremental backup software would copy all of the data that changed since Monday. On Wednesday it will copy everything that changed since Tuesday, and so on.
As you can see, this backup strategy removes the main problem of differential backup – the numerous copies of data (data that is already backed up). Companies with a capable incremental backup software may even benefit further by performing backups more than once a day.
However, incremental backup has its own drawbacks, the main one being the fact that it can affect restore time – data needs to be reconstructed from all of the Incremental backups since the last Full - to be able to perform the restoration process.
There’s always the question of which backup strategy is the best. The fact is, they all have their advantages and downsides, and specific cases may dictate specific backup levels. To make the comparison easier, we can choose several categories and rank each backup levels performance in each of those. This can make it easier to decide, if you are trying to decide the right backup level for you.
Full vs. Incremental vs Differential backup comparison
|Strategy name||Backup starting point||Speed||Storage requirements||Recovery needs|
|Full||Previous full backup||Slow||Very large||Last backup|
|Incremental||Previous full backup||Decent||Big||Last full + last differential|
|Differential||Any previous backup||Fast||Small||Last full + all incremental since the last full|
A lot of companies prefer to save storage space over faster recovery times (although that, of course, depends on the nature of their business, how they use data, and how time-sensitive data recovery may be).Therefore, Incremental backup might be considered to be slightly more popular than Differential backup.
Incremental and differential backups with Bacula Enterprise
Bacula Enterprise offers not only the above backup levels in its solution, but many more too. And when combined with other methods, including advanced deduplication and compression techniques, Bacula is able to enable very specific backup and recovery strategies to an especially wide range of organizations. As a result of this flexibility, combined with an especially high degree of security, Bacula is used by a large number of military and defence organizations all over the world, as well as MSPs, ISVs, E-commerce companies, large data centers and other very demanding IT environments. Contact us to find out more about our customers in your industry.