by davids355 » Thu Feb 05, 2015 5:40 am
UPDATE:
OK so I had 8 concurrent tasks running tonight. each backup usually lasts between 1-3 hours so fair amounts of data being backed up.
All were backing up to the same remote rsync server (running CentOS).
5 backups were going to one physical harddrive and 3 backups to another.
Whilst the backups were running, CPU was sitting at 2-3 percent.
As soon as I saw some "bash -c rsync" commands, on the remote server, processor for that PID was around 30%, I noticed the --delete command there as well - would that be rsync tidying up old backups?? (See screenshot for example).
As soon as the clients moved to the last stage - compiling usage reports, then CPU for that PID went to constant 100% (Screenshot also attached) - looked like each PID doing compile of usage report was using 100% of a single core on the remote backup system.
Interestingly, once I canceled the job on the corresponding client, it did not stop that process from running on the remote server.
Based on that I am going to run all 8 backups together again tomorrow evening, with usage report feature turned off (i think i saw that in settings), in the hope that the demand on remote server will be greatly decreased...
Screenshot showing several processes on remote backup server, created by backup assist jobs:
[img]bloggtech.com/backup%20assist%20remote%20rsync%20command.png[/img]