Re: [rdiff-backup-users] hi all help for newbie

[Andrew K. Bressen]

manu aradhya <address@hidden> writes:
>  hi all iam very new to rdiff-backup   i have huge
> file (databasefile (binary)) which will grow
> regularly. at first i want to take full backup of it,
> then everyday i want to take incremental backups
> (diffs).i tried with rdiff-backup, 

rdiff-backup does this backwards from most other backup programs.

The first time you make a backup, it copies everything
(makes a full backup).

When you make an increment, 
it does not store the diffs that make the old files into the current files.

Instead, it turns the backup into being the same as the current files, 
and stores the diffs to make it into older versions. 

This is good because the most recent versions of files are always
easy to get to. This does not use extra bandwidth because it uses
the rsync algorithm. 

>                 If i want to restore do i need to
> start from this  first time backed file, and then do i
> need to restore diffs on the same file. 

No, you use rdiff-backup to make restores, 
and you tell it what time you want the restore to look like.
It starts with the most current version of the files, 
and then it applies diffs to make it into older versions if 
that is what you need. . 

> i just need a full backup of the source file for the
> first time (this fullbackup is going to written on to
> cdrw).and everyday a incremental backup (diff).how to
> restore once again ? pls somebody suguest me some
> ideas.

rdiff-backup does not really work like this.
it needs to always be able to get to the full backup and change it,
so burning it onto a cd is not the way to go. 

Hope this helps.

  --akb