clone performance of experimental new http client library.

Augie Fackler durin42 at gmail.com
Wed Oct 19 10:39:22 CDT 2011


On Oct 19, 2011, at 10:32 AM, <David.Sedlock at lantiq.com> <David.Sedlock at lantiq.com> wrote:

> Hi all,
> 
> A little more info: The slowdown is not linear. Cloning a 40M repo is 3x slower with ui.usehttp2=True, while cloning a 277M repo is 35x slower.
> 
> ***1.9.2 client with 1.9.2 server, ui.usehttp2=False, 277M repo
> $ hg version
> Mercurial Distributed SCM (version 1.9.2-lantiq-rc1)
> $ hg --time clone -U https://mtstest.lantiq.com/~hgtest/repo/openwrt/core/ --config ui.usehttp2=False
> Time: real 115.310 secs (user 69.310+0.000 sys 5.180+0.000)
> 
> ***1.9.2 client with 1.9.2 server, ui.usehttp2=True, 277M repo
> $ hg version
> Mercurial Distributed SCM (version 1.9.2-lantiq-rc1)
> $ hg --time clone -U https://mtstest.lantiq.com/~hgtest/repo/openwrt/core/ --config ui.usehttp2=True 
> Time: real 4131.950 secs (user 2211.010+0.000 sys 1898.440+0.000)
> 
> It seems as if the cloning goes slower and slower. For example, it starts off pulling the manifests at a good clip and then gets slower and slower till the progress arrow is moving with agonizing slowness by the end. Same for file changes. The progress prediction is always way off.

Can you get me a public repo to test against? I'm happy to spend some time in a profiler and speed things up, but I need a way to test.

> 
> Ouch!
> 
> Regards,
> David
> 
> 
>> -----Original Message-----
>> From: mercurial-devel-bounces at selenic.com [mailto:mercurial-devel-
>> bounces at selenic.com] On Behalf Of Sedlock David (LQKG IT RDS)
>> Sent: Wednesday, October 19, 2011 11:23 AM
>> To: mercurial-devel at selenic.com
>> Subject: clone performance of experimental new http client library.
>> 
>> Hi all,
>> 
>> Did anyone benchmark the peformance of the experimental new http client
>> library? After experiencing significantly longer clone times with 1.9.2, I
>> performed the following benchmark:
>> 
>> ***1.9.2 client with 1.9.2 server, ui.usehttp2=False
>> $ hg version
>> Mercurial Distributed SCM (version 1.9.2-lantiq-rc1)
>> $ hg --time clone -U https://mtstest.lantiq.com/~hgtest/repo/openwrt/luci/ --
>> config ui.usehttp2=False
>> Time: real 21.800 secs (user 9.670+0.000 sys 0.650+0.000)
>> 
>> ***1.9.2 client with 1.9.2 server, ui.usehttp2=True
>> $ hg version
>> Mercurial Distributed SCM (version 1.9.2-lantiq-rc1)
>> $ hg --time clone -U https://mtstest.lantiq.com/~hgtest/repo/openwrt/luci/ --
>> config ui.usehttp2=True
>> Time: real 63.310 secs (user 33.790+0.000 sys 21.170+0.000)
>> 
>> 3x slower.
>> 
>> Regards,
>> David
>> 
>> Lantiq Configuration Management Solutions
>> http://wiki.lantiq.com/display/IT/Configuration+Management
>> __________________________________
>> Lantiq Beteiligungs- GmbH & Co.KG
>> Registered Office: München, Commercial Register: Amtsgericht München, HRA
>> 94167
>> Limited Partner (Kommanditist): Lantiq Intermediate Holdco S.à r.l
>> General Partner (Komplementär): Lantiq Beteiligungs Verwaltungs- GmbH
>> Registered Office: München, Commercial Register: Amtsgericht München HRB
>> 180523
>> Managing Directors (Geschäftsführer): Christian Wolff (CEO), Dr. Klaus Gohlke
>> (CFO) ___________________________________________
>> Important Note: This e-mail may contain trade secrets or privileged,
>> undisclosed or otherwise confidential information. If you have received this
>> e-mail in error, you are hereby notified that any review, copying or
>> distribution of it is strictly prohibited. Please inform us immediately and
>> destroy the original transmittal. Thank you for your cooperation.
>> 
>> 
>> _______________________________________________
>> Mercurial-devel mailing list
>> Mercurial-devel at selenic.com
>> http://selenic.com/mailman/listinfo/mercurial-devel
> _______________________________________________
> Mercurial-devel mailing list
> Mercurial-devel at selenic.com
> http://selenic.com/mailman/listinfo/mercurial-devel



More information about the Mercurial-devel mailing list