What is an image-level backup and how is it different from traditional backup?
Image-level backup is a snapshot. There are three different kinds of snapshots: Copy-on-write (COW), Redirect-on-write (ROW), and clones. COW and ROW really are taking only a copy of the pointers of what the data looked like at that point in time. There is no real copy of the data although COW will create one over time as changes to that original data occur. Clones are complete copies of the data.
Traditional backup refers to creating a full copy of data. Today, backup software typically backs up all of the files on a server including the hidden files one time, then backs up any changes since the previous backup.
The differences between the two technologies show up in restores. A snapshot has to be mounted to recover any of the data. For single file recoveries, which are more than 90% of the restores per Symantec's recent research, traditional backup (also known as file-level backup), is a better data protection solution because only that single file needs to be restored.
For system-level restores, image backups or snapshots are much faster and easier to restore. It's not an either or and they are not mutually exclusive. Many backup software products provide both image and file backup as well as file and image restores.
This was first published in December 2013