In order for a backup plan to truly protect against data loss, it’s essential that the backup policy keeps multiple historical versions of all documents and files.
If your backup process simply consists of duplicating or mirroring data between your computer and an external drive, then you won’t be protected against viruses or file corruption. If your primary hard drive become corrupted, your backup process will simply overwrite your “backups” with corrupted data.
This shows why maintaining a rotating set of point-in-time backup copies is critical. However, there are a number of challenges associated with this approach.
If you’re maintaining a 30 day backup policy for a 1 TB hard drive, you’d need 30 TB worth of backup media (tapes or USB drives). This can be very expensive. That’s why “differential” backup methodologies were invented.
A “differential backup” or “incremental backup” process is simply a more economical approach that aims to eliminate waste by ignoring stale data and only focusing on backing up data that has recently been modified in some way.
- If a folder contains 10 files, and you modify one of those files, that’s the only file that needs to be backed up.
- If a folder contains 10 files, and then you add a new file, then only this new file requires a backup.
It’s really that simple.
(Differential backups get more complicated when you start dealing with database systems, but we’ll cover this in another article.)
When performing incremental backups, there are 2 leading approaches: Block-Level and File Level
File Level Incremental Backup
The example outlined above illustrates the classic File Level Incremental backup. If a file has been added or modified, it copies the entire file — and only that file — to backup.
But this approach has some problems. If there is a very large file, and that file gets appended or slightly modified in some way, then there is a lot of waste. (For example: A log file that is several gigabytes in size, and then a single line is appended to the end of the file)
Another issue arises when large volumes of data require backup, but there is limited network bandwidth to transfer these backups. (This is a common problem for laptop users that rely on certain types of cloud storage)
Block-Level Incremental Backup
In order to reduce the amount of storage required for backup, and also to help reduce the time required for backups, Block-Level Incremental methodologies were developed.
Block-level incremental backups will analyze a file or document in order to establish which portions have been modified since the last backup. Then, it will only copy over the specific data blocks that have been modified, instead of copying the entire file.
Block-Level Incremental backups are significantly more efficient than File-Level backups, and are ideal whenever time or storage costs are a concern.
To learn more aboutÂ Block Level and File levelÂ incremental backups, we’ve included this helpful video.