For Windows servers with large shares containing millions of files, backups are very slow. It could be much faster to backup the raw disks, as there wouldn't be the overhead of dealing with opening and closing millions of files.
However we'd still want to be able to do single file restore. It should be possible to analyze the raw disk backup after the backup and construct the file list, and also to be able to mount the raw backups so that single files can be restored. The analysis after the backup should not take longer than a file level backup (and ideally faster) otherwise there is no gain.
This may be difficult to do for incremental backups, but if the raw disk backup speed is fast enough, it may be possible to do full backups all the time (which has advantages for restores).
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.