From mboxrd@z Thu Jan 1 00:00:00 1970 From: Nick Dokos Subject: Re: git repository via http broken Date: Fri, 29 Jun 2012 13:50:02 -0400 Message-ID: <17195.1340992202@alphaville> References: <87pq8mm7ge.fsf@norang.ca> <87lijam769.fsf@norang.ca> <29344.1340748313@alphaville> <87k3yqvz9y.fsf@riotcharge.dunsmor.com> Reply-To: nicholas.dokos@hp.com Return-path: Received: from eggs.gnu.org ([208.118.235.92]:42200) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1SkfS5-0004zi-8f for emacs-orgmode@gnu.org; Fri, 29 Jun 2012 13:57:54 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1SkfS3-00023e-AK for emacs-orgmode@gnu.org; Fri, 29 Jun 2012 13:57:52 -0400 Received: from g6t0185.atlanta.hp.com ([15.193.32.62]:3238) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1SkfS3-00022a-4t for emacs-orgmode@gnu.org; Fri, 29 Jun 2012 13:57:51 -0400 In-Reply-To: Message from Jason Dunsmore of "Fri, 29 Jun 2012 11:49:45 CDT." <87k3yqvz9y.fsf@riotcharge.dunsmor.com> List-Id: "General discussions about Org-mode." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org Sender: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org To: Jason Dunsmore Cc: Bernt Hansen , emacs-orgmode@gnu.org Jason Dunsmore wrote: > On Tue, Jun 26 2012, Nick Dokos wrote: > > > PS BTW, I know that git gc is recommended on local repos, but is that > > applicable to the repo on orgmode.org? and would that help with the > > cloning time? 2.5 mins seems long to me. If not gc, is there > > something else that would help? > > Hi Nick, > > It looks like "git gc" was needed on the server. > > Before git gc on server: > > $ time git clone git://orgmode.org/org-mode.git > Cloning into org-mode... > remote: Counting objects: 62414, done. <- Server disk I/O bottleneck > remote: Compressing objects: 100% (18532/18532), done. <- Server CPU bottleneck > remote: Total 62414 (delta 49455), reused 55032 (delta 43761) > Receiving objects: 100% (62414/62414), 55.46 MiB | 1.16 MiB/s, done. <- Network bottleneck > Resolving deltas: 100% (49455/49455), done. <- Local CPU bottleneck > > real2m44.757s > user0m11.269s > sys0m0.740s > > After git gc on server: > > $ time git clone git://orgmode.org/org-mode.git > Cloning into org-mode... > remote: Counting objects: 62414, done. > remote: Compressing objects: 100% (12836/12836), done. > remote: Total 62414 (delta 49457), reused 62414 (delta 49457) > Receiving objects: 100% (62414/62414), 55.41 MiB | 1.12 MiB/s, done. > Resolving deltas: 100% (49457/49457), done. > > real0m59.071s > user0m11.305s > sys0m0.668s > That's a pretty good speed-up! I tried the same with http just now and got: $ time git clone http://orgmode.org/org-mode.git Cloning into org-mode... real 3m30.619s user 0m10.970s sys 0m0.750s so it improved on the 5 mins I was getting before by roughly 35%. My git time was pretty bad though (4.5 mins): I suspect clogging on the link to my ISP, so both of these numbers are probably suspect. I'll try again later tonight and see what I get. > I've set up a weekly cron job to run "git gc" on the org-mode repo. > Great! Thanks, Nick