craigni wrote: <snip : I'm getting enamored of rsync to the point where I think that's the : backup strategy I'd like to use <snip> Since you are interested in using rsync for backups, you might profit by having a look at "rnsapshot" which uses rsync at its heart but with a nice way to configure what and how often you want to backed up. <http://www.rsnapshot.org/> An alternative to this is another rsync-based backup system: <http://edseek.com/~jasonb/articles/dirvish_backup/introduction.html> These both use the hardlink facilities of rsync to only backup what's changed, yet keep your entire directory structure intact for each backup (by using multiple hard links to the same data). There exist files for which rsync, as used above, is not a good choice. These are typically large files to which you make frequent small changes. The above packages will make a new copy of the _entire_ file each time you make a change. If the file is, say, 20GB, well, you see the problem. The virtual disk files used by VMware are examples of such files. Every time I merely open a "Word Doc" M$ sees fit to update the doc file. The VMware virtual disk therefore sees a (tiny) change. In this case there's <http://www.nongnu.org/rdiff-backup/>. I've not used this so can't comment first-hand on it, but it is supposed to solve the "big file"/"small changes" problem. Dean
