Re: [rdiff-backup-users] CPU overhead on local and remote ends

[Andrew Ferguson]

On May 9, 2008, at 7:55 AM, Jordi Mallach wrote:
> I'd like to do parallel processes, so I ask: if I execute rdiff-backup
> on the remote servers instead from the backup servers, would that be
> better in terms of CPU usage on the backup server, or do the diff
> calculations happen on the remote end?

Hi Jordi,

rdiff-backup uses the common rsync algorithm:

http://samba.anu.edu.au/rsync/tech_report/
http://en.wikipedia.org/wiki/Rsync

which calculates checksums on *both* ends of the connection to  
minimize network bandwidth.


Andrew