On 13.05.2010 13:24, N. Yaakov Ziskind wrote:> I have 5 or 6 :-( different copies of a filesystem on various Linux
> boxen, all backups taken at different times, with different exclusions,
> and squirreled away. Now's the time to clean up my attic. I'd like
to
> merge them all into one big filesystem. When there are different copies
> of the same files, I'd like to keep the newest; I don't know what
else
> to do. My plan (assuming I'm going to retain 'path0' as the
surviving
> filesystem) goes like this:
>
> rsync -azPv -u path-to-system-1 path0
> rsync -azPv -u path-to-system-2 path0
> rsync -azPv -u path-to-system-3 path0
> rsync -azPv -u path-to-system-4 path0
> rsync -azPv -u path-to-system-5 path0
>
> and then rm -rf "path-to-system-?", and hope for the best.
>
> What kind of disaster can I look forward to with this approach, and is
> there a better one?
It depends on WHAT and HOW MUCH.
When i had the same situation a few years back for my /usr/local/bin i
choose a "master" machine and made that directory into a subversion
repository and then on each of the other 2 machines i decided
file-by-file what what was the "right(tm)" version to commit, revert,
intergrate or delete.
For Backup-Purposes you could keep the backups and use a filedup-utility
to hardlink identical files to reduce the space needed. That what you
can at least always look back. And delete is a few years later. ;-)
Bis denn
--
Real Programmers consider "what you see is what you get" to be just as
bad a concept in Text Editors as it is in women. No, the Real Programmer
wants a "you asked for it, you got it" text editor -- complicated,
cryptic, powerful, unforgiving, dangerous.