Just in case anyone cares, I performed a set of tests on this system and things seem to work OK with --ignore-times still on.
A full backup of a 750GB backup set took 14 hours on the local network, but only transfered about 4.3GB of raw data in and 1.1GB out. There we only about 2GB in changed files since the last backup. I think we can move this off-site without fear of huge bandwidth bills. Nick Nick Webb wrote: > Hi all - > > Likely a stupid question, but is it safe to remove the --ignore-times > rsync option in the BackupPC config.pl file? When I've done rsync > backups before with simple scripts, I've never used that option, and > have yet to land in hot water because of it. While in an ideal world > we'd check the file contents every time, it seems pretty unlikely to > bear any fruity 99.999% of the time. > > The reason I'm considering axing this option is that I'm setting up a > remote backup server backing up about 1TB of data. I'm concerned that > running a full backup in the future will take days to complete, when > really few files changed (initial full backup was done on the local > network). The remote network link is about 1Mbit/sec, and there is > other data going over this link, so really we'll have about 500Kbit for > backup, or less. > > This also ties in with the discussion "Full backup bandwidth reduction". > If BackupPC really is bandwidth efficient during the full backup, I > could leave --ignore-times on, but so far I'm not convinced. I don't > have any solid data yet, though. I guess we'll see soon. > > I'm also concerned about disk utilization as the backup server has only > about 1TB of disk space and the backup set compressed is around 500GB. > If subsequent full backups will take another 500GB, even temporarily, > we'll be in trouble. Sounds like this shouldn't be the case, but I'm > not convinced yet on this either based on what I've seen so far. > > I've read through the config.pl file and there is an option (marked > useless) for doing "filled" incrementals, which would seem to do what we > want. Just do one full backup, then incrementals after that... Of > course if all the files end up matching what's in the pool on a full > backup, everything should be hunky dory without using filled incrementals. > > Thanks in advance for your thoughts! > > > Thanks, > Nick > ------------------------------------------------------------------------- SF.Net email is sponsored by: The Future of Linux Business White Paper from Novell. From the desktop to the data center, Linux is going mainstream. Let it simplify your IT future. http://altfarm.mediaplex.com/ad/ck/8857-50307-18918-4 _______________________________________________ BackupPC-users mailing list [email protected] List: https://lists.sourceforge.net/lists/listinfo/backuppc-users Wiki: http://backuppc.wiki.sourceforge.net Project: http://backuppc.sourceforge.net/
