Obviously... Data changes more often than system files. Have your system drive be only system files. Have a separate data/program partition.
Back up your your system files on a semi regular basis. Back up your data every time. No matter how you end up doing it, this will cut down on backup sizes and time spent. I've had some limited experience with g4u as drew brought up. It's a pretty neat little tool! Seems to work well in the few times I've used it. -Nye Walker On Fri, May 29, 2009 at 1:29 AM, Chris Daniel <[email protected]> wrote: > You could mount the filesystems read-only using ntfs-3g (or there > might even be something in kernel, or ntfs-3g might be in kernel -- I > don't keep track anymore) and then rsync. Assuming, of course, Win 7 > is still using NTFS. Like Chris is saying, the disk imaging is going > to be your biggest time sucker, not booting. DSL sounds good to me for > a minimal install, however. > > On Fri, May 29, 2009 at 1:14 AM, chris (fool) mccraw <[email protected]> > wrote: > > On Fri, May 29, 2009 at 00:23, Keith Lofstrom <[email protected]> wrote: > > > > I'm that jerk who questions your methods instead of answering your > question. > > > >> D.S.Linux runs dd to clone the main hard > >> drive to an identical hard drive in an external SATA cradle, then > >> shuts down. > > > > dd is hardly the fastest way to copy data. i see what you're going > > for, though--obviously rsync or similar "don't copy already-same > > areas" tool won't necessarily create a bootable copy of windows. i > > didn't find any tools like dd+xdelta out there but there must be a > > more efficient tool...does something like ghost (non free, but if > > you're dealing with windows anyway...) do that? i can't imagine that > > 95% of the disk changes each usage, so it seems like an area of this > > process rich for optimization. i wonder if you have already optimized > > partition size down to the bare minimum and just dd that partition? > > seems like optimizing boot time when it must be such a tiny part of > > the time taken to make the copy is ignoring amdahl's argument that one > > should optimize the slowest thing first rather than the easiest... > > > > sorry i don't have an actual answer for you :( > > > > luck++; > > _______________________________________________ > > PLUG mailing list > > [email protected] > > http://lists.pdxlinux.org/mailman/listinfo/plug > > > > > > -- > Chris Daniel > _______________________________________________ > PLUG mailing list > [email protected] > http://lists.pdxlinux.org/mailman/listinfo/plug > -- Nye Walker http://www.netnye.com _______________________________________________ PLUG mailing list [email protected] http://lists.pdxlinux.org/mailman/listinfo/plug
