How to prevent data corruption.

[Andrea Riciputi]

Hi all,
I'm setting a script to backup my system using rsync. After reading  
the manual and googling the web a lot I'm wondering how rsync can  
protect my data from any potential corruption of the source HD.

Suppose that a file on the source HD has been corrupted (because of a  
device failure), how can I avoid that the corrupted version of the  
file will override the backup version already present on the backup  
HD? Can the --checksum option be used for this purpose even if I'm  
doing a local backup?

Any other suggestion or backup strategy are also wellcome. (Obviously  
I know that backup strategies can be very complex, but what I'm  
looking for here is a fairly good solution for a home-based system).

Thanks,
  Andrea.