[csw-maintainers] experimental: subversion --> git, mercurial

Sebastian Kayser skayser at opencsw.org
Sun Jun 26 16:47:57 CEST 2011


* rupert THURNER <rupert at opencsw.org> wrote:
>    to get a feeling for the question how to get a full checkout, i created:
>    * [5]https://github.com/opencsw/pkg  (submodules)
>    * [6]https://github.com/opencsw/pkg-all (full repo)

Cool, thanks alot for the experiment!

>    the pkg-all feels quite heavy. creating it took nearly 9 hours, its 100mb
>    big, takes more than 5 minutes to push and clone, see below. and, it
>    cannot be cloned in parts, like the subversion tree before. and, it is not
>    open to have partial migration.
>    $ time git svn clone
>    [7]https://gar.svn.sourceforge.net/svnroot/gar/csw/mgar/pkg pkg-all
>    ...
>    real    509m37.383s
>    user    109m23.710s
>    sys     378m14.304s
>
>    rupert at login:~/git-experiment/pkg-all
>    $ time git push origin master
>    Counting objects: 87829, done.
>    Delta compression using up to 24 threads.
>    Compressing objects: 100% (51647/51647), done.
>    Writing objects: 100% (87829/87829), 95.34 MiB | 352 KiB/s, done.
>    Total 87829 (delta 31596), reused 85545 (delta 29312)
>    To git at github.com:opencsw/pkg-all.git
>     * [new branch]      master -> master
>    real    5m6.088s
>    user    0m56.358s
>    sys     0m6.002s

The operations above are one off, right? To add some further figures, a git
clone of the full repo takes ~2mins for me, git log . in a package directory
~12 seconds.

$ time git clone git://github.com/opencsw/pkg-all.git
remote: Counting objects: 87829, done.
remote: Compressing objects: 100% (49363/49363), done.
Indexing 87829 objects.
remote: Total 87829 (delta 31596), reused 87829 (delta 31596)
 100% (87829/87829) done
Resolving 31596 deltas.
 100% (31596/31596) done
Checking files out...
 100% (12273/12273) done

real    1m56.048s
user    0m16.952s
sys     0m3.641s

$ pwd
/home/ska/tmp/gittest/pkg-all/lang-python/sqlobject/trunk
$ time git log . >/dev/null

real    0m11.587s
user    0m11.532s
sys     0m0.052s

Sebastian


More information about the maintainers mailing list