Re: [rdiff-backup-users] incremental or differential backup

[Andrew Ferguson]

On Jan 16, 2009, at 7:00 AM, Michael Biebl wrote:
> 2.) If one of the rdiffs goes corrupt (e.g. via a bad sector), all my
> older backups are broken.


Every 10 incremental diffs, rdiff-backup stores another snapshot of  
the file.

So, let's say you have 22 versions of your file. Then, in the rdiff- 
backup-data, the two most recent backups would be stored as reverse- 
diff increments. Next, there would be a snapshot. Then, nine reverse- 
diff increments. Then, another snapshot, then the first nine reverse- 
diff increments.

During the restore, rdiff-backup finds the oldest snapshot at least as  
recent as the desired backup time (it could be the current mirror, or  
one of these snapshots). Then, it will apply any reverse-diffs  
necessary to recreate the desired version.

These webpage can provide some guidance on the terms:
http://www.nongnu.org/rdiff-backup/format.html


> 3.) I can't throw away rdiffs, say I want to create daily snapshots,
> keep 30 of them, then keep monthly snapshots, and so on (basically
> what tools like rsnaphost provide).


That is because there is no support from the librsync folks for  
merging the diffs directly. Without such support, merging diffs would  
be very disk intensive (create the two complete files, then create the  
third diff).

See http://wiki.rdiff-backup.org/wiki/index.php/DeleteIntermediate


Andrew