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

[Dominic]

Michael Biebl wrote:
> 2009/1/16 Dominic <address@hidden>:
>   
> > The reason I use rsync for the secondary backup is that I just want a 
> > mirror
> > of the data on the first (rdiff-backup) backup machine - including (and
> > especially) all the rdiff-backup archives. Doing an rdiff-backup of an
> > rdiff-backup archive would seem too much of a good thing. Or maybe I just
> > can't get my head around it.
> >     
> Ah, I didn't mean it that way
> What I had in mind was something like:
>
> rdiff-backup /foo /backup/
> rdiff-backup /foo remote-server::/backup/
>
> (both triggered via cron, e.g.)
>   
in that case rdiff-backup would be (IMO) the best solution for both backups.

The advantages that I can see to rsync are relative simplicity (though 
maybe not once you are using rsnapshot on top of it) and the comfort 
factor that comes from knowing "one million administrators can't be 
wrong". Rdiff-backup is certainly less widely used, but it's not on the 
bleeding edge either.

For my money (and it doesn't take any!) the only thing that rdiff-backup 
lacks as a backup solution is data security against a malicious 
administrator. There is a workaround involving encfs over sshfs, but 
(not having tried it) I expect there is a heavy performance penalty. It 
also has no built-in data security against theft of the backup machine 
or disk, but this can be covered I think by using an encrypted 
filesystem. AFAIK the best secure open-source data-backup solution is 
Box Backup, which uses reverse incremental diffs just like rdiff-backup 
(but I haven't used it).

Dominic