Why is Block Level Backup Best?
Blueshift approaches data backup and recovery from a slightly different angle than most traditional backup companies. Traditional backup technology is typically based on the concept of incremental backups which means your backups are performed at the file level, Blueshift DataProtect enables you to perform backups at the block level.
So you may be thinking what’s so wrong with doing file level backups? Well the first thing you need to understand is that file-level backups consume hugs amounts of compute resources. This means that the backup application itself is actually competing with critical compute resources on the very system it’s supposed to be backing up, not a very good idea.
Another advantage of block-level backups is that they bypass the file system so there is no penalty on backup performance for having a large number of files. The backup application never looks at files and doesn’t really care how many files there are on a server. Block-level backups start by taking a snapshot of the live running volume. The block-level data is then read from the snapshot, rather than the actual disk. This allows the backup to run in harmony with your file system, rather than taxing your servers and causing performance issues.
Remember, the file system is responsible for keeping track of the tree or hierarchy of files. In doing so, it stores files in neat little fixed size blocks on the disk. Unfortunately, files can be located on blocks that are scattered all across the disk. Because of this, backup applications that read data at the file level are forced to jump all around the disk and traverse the directory tree just to find all of the files eligible for backup. This tends to be a very fragmented, convoluted process that causes the disk to spend as much time, if not more, searching for the correct information instead of actually reading the data.
Block-level backups are extremely efficient when compared to file-level backups. Since they are able to read blocks in the order that they actually appear on the disk, rather than the order that they appear in files, these solutions are able to save a ton of time jumping back and forth across the disk.