Hi,

Tino Schwarze wrote on 2008-08-28 23:33:40 +0200 [Re: [BackupPC-users] Full vs. 
Inc]:
> On Thu, Aug 28, 2008 at 11:24:16PM +0200, Christian Völker wrote:
> > [...]
> > one of my questions regards the difference between Full vs. Inc Backups.
> > 
> > Does the full backup copy all files, not checking if they already exist
> > on the backup target? But where is the sense then in using rsync?
> 
> The following applies to rsync only!
> 
> The difference between a full and an incremental rsync backup is that
> rsync will check the whole file checksum during full backup (to guard
> against pool file corruption

... and undetected alteration of the file on the client (same stats including
size and timestamp - probably rare, but not impossible) ...

> ). It will not transfer the whole file. The
> incremental backup will transfer all differences since the last full or
> incremental (depending on incremental level).
> 
> > Or is it the way the backup is stored?
> 
> A full backup will have a fully populated directory structure on the
> server (with hardlinks to the common pool) while an incremental backup
> only has the directories and any changed files (again hardlinked to the
> common pool).
> 
> To sum it up: There are subtle differences in transfer

Since every file needs to be completely read on the client (and the server
too, if you're not using checksum seeds) for a full backup, the difference
in backup duration may be less than subtle though ...

> (only for safety reasons) and major differences in storage.

How "major" the differences in storage are probably depends on your data set.
The important point is that in *both* cases, the amount of *file content*
stored is identical due to pooling(*) (*directory* sizes differ). Your view
of the backup via the web interface or BackupPC_tarCreate et al. is identical
too, meaning you don't have to "fill in" the missing files in an incremental
backup in any case. BackupPC does that for you automatically.

Another important difference is in the dependencies. An incremental backup
depends on the preceeding backup of the next numerically lower level (meaning
a full backup in the simple case). As long as the incremental backup remains
valid, all backups it depends on need to be kept as well, even if they would
otherwise expire. This also means that constructing the view of the backup
(needed for making the next backup as well as viewing the backup in the web
interface or restoring files from it) becomes more expensive with increasing
incremental level. Probably nothing to worry about for level 1 incrementals
though.

Regards,
Holger


(*) in almost all cases that is - I guess exceeding $Conf{HardLinkMax} for a
    file would result in an additional copy of this file being made, so a full
    backup could in very rare cases occupy more space in terms of file content
    than an incremental would have. If you don't understand what this means,
    just ignore it. It's not important for practical purposes.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
BackupPC-users mailing list
BackupPC-users AT lists.sourceforge DOT net
List:    https://lists.sourceforge.net/lists/listinfo/backuppc-users
Wiki:    http://backuppc.wiki.sourceforge.net
Project: http://backuppc.sourceforge.net/