[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Subversion authentication via SASL GSSAPI and likewise open

From: slaventii <xumuku_at_gmail.com>
Date: Wed, 1 Aug 2012 08:04:20 +0300

My last test results :
SVN Server:
Ubuntu 12.04
svn, version 1.7.5 (r1336830) compiled Jul 19 2012, 21:53:29
SVN Client:
Windows Server 2003
svn, version 1.7.5 (r1336830) compiled May 15 2012, 12:29:08
LAN 100 Mbit/s

co svn:// ~ 37m 44 sec
co https:// ~ 40m 04sec
co svn vs https perf, % ~ > 6%

up svn:// ~ 0m 35sec
up https:// ~ 0m 19 sec
up svn vs https perf, % ~ > 23%

svn client 1.5.4 with svn server 1.7.5 with same servers:
co svn:// - 22m 30 sec

cat /etc/apache2/mods-enabled/dav_svn.conf
<IfModule dav_svn_module>
      # Enable a 1 Gb Subversion data cache for both fulltext and deltas.
      SVNInMemoryCacheSize 1048576
      SVNCacheTextDeltas On
      SVNCacheFullTexts On
      SVNCompressionLevel 0
      SVNAdvertiseV2Protocol On
</IfModule>

cat /etc/xinetd.d/svnserve
       server_args = -i -r /var/svn/repos
--log-file=/var/log/svn/svn.log --memory-cache-size=1024
--cache-txdeltas=yes --cache-fulltexts=yes --compression=0

In my result I see that co operation decrised very high when I used new client.
Also co operation time has no so big difference between svn and http protocols,
But up operation is 23 % slower via http.

Maybe this is because of "SVNCompressionLevel 0" ?
"For example, on a local area network (LAN) with 1-Gigabit
connections, it might not make sense to have the server compress its
network transmissions (which also forces the clients to decompress
them), as the network itself is so fast that users won't really
benefit from the smaller overall network payload. On the other hand,
servers which are accessed primarily by clients with low-bandwidth
connections would be doing those clients a favor by minimizing the
overall size of its network communications."
But my LAN is 100 Mbit/s

On Tue, Jul 31, 2012 at 9:48 PM, Johan Corveleyn <jcorvel_at_gmail.com> wrote:
> On Tue, Jul 31, 2012 at 6:59 PM, slaventii <xumuku_at_gmail.com> wrote:
>> Want to share some of my tests results:
>> /trunk - 1.80 GB (1,941,844,940 bytes) - 148,114 Files; 52,519 Folders
>> --------------------------------------------------------------------------------
>> Old Hardware Server - Ubuntu 9.04:
>> svn, version 1.5.4 (r33841) compiled Aug 7 2009, 01:44:11
>> co svn:// ~ 23m 32sec
>> co https:// ~ 30m 10sec
>> svn vs https perf, % ~ > 28%
>>
>> up svn:// ~ 3m 22sec
>> up https:// ~ 5m 04sec
>> svn vs https perf, % > 50%
>> --------------------------------------------------------------------------------
>> New Hardware Server - Ubuntu 12.04:
>> svn, version 1.6.17 (r1128011) compiled Dec 17 2011, 16:12:52
>> co svn:// ~ 18m 30sec
>> co https:// ~ 22m 47sec
>> svn vs https perf, % ~ > 23%
>>
>> up svn:// ~ 2m 06sec
>> up https:// ~ 2m 35sec
>> svn vs https perf, % ~ > 23%
>> --------------------------------------------------------------------------------
>> New Hardware Server - Ubuntu 12.04:
>> svn, version 1.7.5 (r1336830) compiled Jul 19 2012, 21:53:29
>> co svn:// ~ 22m 50sec
>> co https:// ~ 24m 00sec
>> svn vs https perf, % ~ > 5%
>>
>> up svn:// ~ 2m 38sec
>> up https:// ~ 2m 28 sec
>> svn vs https perf, % ~ < 7%
>> --------------------------------------------------------------------------------
>> I can see that in general performance didn't increase so much.
>> But difference between svn and http decreased.
>
> What kind of network connection between server and client?
>
> The svn version numbers you report, are those the versions of the
> server, or of the client, or both?
>
> --
> Johan
Received on 2012-08-01 07:04:56 CEST

This is an archived mail posted to the Subversion Users mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.