At the Ottawa Linux Symposium, I heard talk of a proof-of-concept version
of diff rewritten to use the rsync algorithms to produce diffs across a
network. Does such a beast still/actually exist?
Background:
I'm putting together a computer lab for use in operating systems classes,
and the school has decided that the best way to do this is through the use
of VMWare, where each student gets a VMWare disk image to do their work
in. (Images average around 400MB)
Right now, each student has to pick a box as theirs for the semester, and
if another student comes along and trashes that box and loses their image,
they're up the creek. I'd like to implement a way for the images to be
stored server-side for central backups and so that students can go from
machine to machine.
The basic methodology here would involve having each box store a local
copy of each default image (RH6.2, Win95, etc.) on disk, and have the
server store a diff for each student. Flow would be similar to this:
Student logs in to box:
Box determines the target OS based on conf files
Box contacts server and downloads checksum for the default image
for that OS, and confirms that its local default image
matches.
If the checksum fails, rsync to server to ensure its default image
is correct. (Maybe eliminate the previous step with this one)
Download the differences between the student's latest working copy
and the default image, and apply the differences to a local
copy of the default image.
Launch VMWare on the temporary image, student plays, saves to the image.
Student exits VMWare to log out:
Box reconfirms checksum of local image from server.
Box builds diff file between local default image and local temporary
image, and uploads that diff file to the server for archival.
Does anyone have any thoughts on what I'm trying to accomplish? Would
there be better tools for doing this than a diffing tool based on rsync?
Thanks,
--
Rob Russell Professional Services Senior Systems Analyst
Nitrosoft Linux [EMAIL PROTECTED] 1-877-OK-LINUX x322
