Les Mikesell writes: > On Fri, 2006-01-20 at 16:44, Dan Pritts wrote: > > On Fri, Jan 20, 2006 at 03:53:33PM -0600, Les Mikesell wrote: > > > > > > I'd expect to see quite a lot of temp file activity that would > > > result in changes to unused space on the live system. It would > > > > Yeah, you're right - i think of this as all being rsync based > > but it sort of is and sort of isn't. Files get transferred > > to the store disk and then get pooled, rather than using > > rsync --link-dest. > > Even a normal rsync transfer of anything with changes builds > a new copy of the file and replaces the old only when the > new temp copy is complete. Hmmm... I guess that means the > partition image scheme needs disk space for a 2nd copy > of the file at the remote side.
Rsync's --inplace option solves this issue: it doesn't use a temp file. The restriction this imposes is that no blocks will be matched earlier in the file, since with --inplace the earlier blocks might already be changed. Also, --inplace will of course leave the target in an undefined state if the transfer is interrupted. For a partition sync the matching limitation wouldn't be significant since most matching blocks will be at the same offset. The notion of using rsync on the raw device is intriguing, provided they are offline or are a frozen snapshot. Craig ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642 _______________________________________________ BackupPC-users mailing list BackupPC-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/backuppc-users http://backuppc.sourceforge.net/
