Il 08/04/2012 13:11, Tim Sutton ha scritto:
Yes - I almost have all the tests passing for QGIS now. After that I
want to set up automated testing so that as soon as something gets
broken we are all made aware of it. It would be nice to have a place
to do that.
Should we go and ask?
On Tue, Apr 10, 2012 at 10:54:13PM +0200, Jürgen E. Fischer wrote:
Hi Tim,
On Sun, 08. Apr 2012 at 13:11:22 +0200, Tim Sutton wrote:
Yes - I almost have all the tests passing for QGIS now. After that I
want to set up automated testing so that as soon as something gets
broken we are all
Hi Radim
Awesome, thanks a lot!
It is very interesting to see the performance relevant changes and it
certainly helps to avoid performance regressions.
Regarding the hosting, I have to check if it is possible to run the
application on a server at Sourcepole.
Regards,
Marco
On 06.04.2012
Some links and notes for those knowing nothing about osgeo
infrastructure (like me):
http://wiki.osgeo.org/wiki/Infrastructure_Transition_Plan_2010
http://wiki.osgeo.org/wiki/AdhocVM
http://osuosl.org/
http://www.linux-kvm.org
If I got it, osgeo3 and osgeo4 are hosted in OSU OSL using KVM
Hi Tim,
On Sun, 08. Apr 2012 at 13:11:22 +0200, Tim Sutton wrote:
Yes - I almost have all the tests passing for QGIS now. After that I
want to set up automated testing so that as soon as something gets
broken we are all made aware of it. It would be nice to have a place
to do that.
BTW every
Il 08/04/2012 23:50, Jürgen E. Fischer ha scritto:
What server are you talking about? AFAIK there in no osgeo4w build server.
Oh, I see. So, are you compiling on your machine, and uploading on osgeo?
Huh?
Well, 9 h of a daily job seems to me a heavy task.
All the best.
--
Paolo Cavallini -
Hi Paolo,
On Mon, 09. Apr 2012 at 08:50:17 +0200, Paolo Cavallini wrote:
What server are you talking about? AFAIK there in no osgeo4w build server.
Oh, I see. So, are you compiling on your machine, and uploading on osgeo?
osgeo4w, yes.
Huh?
Well, 9 h of a daily job seems to me a
Hi
On Sat, Apr 7, 2012 at 9:08 AM, Paolo Cavallini cavall...@faunalia.it wrote:
Il 06/04/2012 13:00, Radim Blazek ha scritto:
To get comparable results, the benchmark must be run on a computer
with no or little other load (Linux kernel counts process switching
overhead in process time) and
Hi
Top posting here to pick up the original thread:
Radim you mentioned QPixmap vs QImage. I was looking at why one of the
unit tests is doing wierd rendering today and happened to read this:
Performance
QPainter is a rich framework that allows developers to do a great
variety of graphical
Il 08/04/2012 12:01, Tim Sutton ha scritto:
This is the type of info that needs to go in the QGIS -Governance
docs. But due to all the consolidation work there is only now:
- qgis.org server hosted by OSGEO
- android.qgis.org hosting donated by me
Thanks Tim for clarifying. Does this
Hi
On Sun, Apr 8, 2012 at 12:46 PM, Paolo Cavallini cavall...@faunalia.it wrote:
Il 08/04/2012 12:01, Tim Sutton ha scritto:
This is the type of info that needs to go in the QGIS -Governance
docs. But due to all the consolidation work there is only now:
- qgis.org server hosted by OSGEO
-
Hi Paolo,
On Sun, 08. Apr 2012 at 12:46:18 +0200, Paolo Cavallini wrote:
Does this include the osgeo4w build server?
What server are you talking about? AFAIK there in no osgeo4w build server.
This seems the most stressed part of the infrastructure now.
Huh?
Jürgen
--
Jürgen E. Fischer
Il 06/04/2012 13:00, Radim Blazek ha scritto:
To get comparable results, the benchmark must be run on a computer
with no or little other load (Linux kernel counts process switching
overhead in process time) and CPU parameters must not change
frequently (all the data collected are relative).
The
Hi Paolo,
On Sat, 07. Apr 2012 at 09:08:41 +0200, Paolo Cavallini wrote:
BTW: can we have an overview of the qgis servers?
Are there more servers than the qgis.org vm?
Jürgen
--
Jürgen E. Fischer norBIT GmbH Tel. +49-4931-918175-31
Dipl.-Inf. (FH) Rheinstraße
On Fri, Dec 9, 2011 at 2:48 PM, Marco Hugentobler
marco.hugentob...@sourcepole.ch dream
A real killer feature would be to have an application that takes a project
and a source tree, runs all the revisions in master branch and outputs the
times for the revisions in a csv file.
/dream
Live
Il 06/04/2012 08:38, Radim Blazek ha scritto:
Live benchmarks: http://www.mpasol.it/blazek/qbench/ .
Currently it is updated every 20 min. Unfortunately I don't have any
server which could be dedicated to it permanently. If anybody can
offer a free server, I can migrate it.
Great job Radim,
On Fri, Apr 6, 2012 at 9:21 AM, Paolo Cavallini cavall...@faunalia.it wrote:
Il 06/04/2012 08:38, Radim Blazek ha scritto:
Live benchmarks: http://www.mpasol.it/blazek/qbench/ .
Currently it is updated every 20 min. Unfortunately I don't have any
server which could be dedicated to it
Hi Radim,
On Fri, 06. Apr 2012 at 13:00:59 +0200, Radim Blazek wrote:
The requirement would be a dedicated time (1-2h per day should be
sufficient) on a server with 2 CPU/cores (the second for database in
case of db providers).
Just so that you know: All the nightly builds on qgis.org
On Fri, Dec 9, 2011 at 2:48 PM, Marco Hugentobler
marco.hugentob...@sourcepole.ch
dream
A real killer feature would be to have an application that takes a project
and a source tree, runs all the revisions in master branch and outputs the
times for the revisions in a csv file.
/dream
On Wed, Dec 7, 2011 at 6:12 PM, Mayeul Kauffmann
mayeul.kauffm...@free.fr wrote:
Hi,
You may try to use http://sourceforge.net/projects/osm2postgresql/
to test with huge OSM datasets.
I'm doing this daily on continental extracts (and planning to do this on
planet.osm soon). (Do NOT try to
Hi Radim
Wow, there are a lot of performance critical changes, even with the
postgresql layers. Thank you very much for the useful qgis_bench, it is
now much easier to make performance tests when implementing a new feature.
dream
A real killer feature would be to have an application that
Le lundi 28 novembre 2011 à 19:49 +0100, Radim Blazek a écrit :
My next plans are:
1) add more projects (more data types and rendering modes) to my
benchmark suite
Hi,
You may try to use http://sourceforge.net/projects/osm2postgresql/
to test with huge OSM datasets.
I'm doing this daily on
On Wed, Nov 30, 2011 at 7:49 AM, Paolo Cavallini cavall...@faunalia.it wrote:
Preparing projects for qgis_bench application, I noticed that qgis
renders the same project much slower than qgis_bench. I found that
QgsMapCanvasMap was using by default QPixmap, not QImage, which is 7
times slower!
These benchmarks are VERY useful expecially for QGIS Server, that
greatly suffers the performance problems.
IMHO:
- it would be easy to setup a test engine for QGIS Server, and keep
updated a table of results, so we could spot promptly any change in
performances
- anyway, QGIS Server
qgis-developer-boun...@lists.osgeo.org scritti il 30/11/2011 11.24.23
IMHO:
- it would be easy to setup a test engine for QGIS Server, and keep
updated a table of results, so we could spot promptly any change in
performances
- anyway, QGIS Server should be used as tiled, in any condition
Hi Luca
Benchmarks are very specific, it's impossible to create general
statements just from the slides without knowing the test arrangement
very well.
If you are referring to the foss4g benchmark this year, the first
comparison test was mainly about the performance of near empty requests
I can confirm that performance in a production environment is ok - see
examples at http://gis.uster.ch/
Performance can always be improved, but one also has to keep in mind
that the other two projects (UMN and Geoserver) are specifically
programmed as a map server whereas QGIS server was
I can confirm that performance in a production environment is ok - see
examples at http://gis.uster.ch/
And in any case, serious performance issues can always be avoided on the
web using tiling and caching.
More serious are the problems with desktop, so I think Radim is right in
focusing there.
Il 28/11/2011 19:49, Radim Blazek ha scritto:
Preparing projects for qgis_bench application, I noticed that qgis
renders the same project much slower than qgis_bench. I found that
QgsMapCanvasMap was using by default QPixmap, not QImage, which is 7
times slower! To be precise, QGIS was using
qgis-developer-boun...@lists.osgeo.org scritti il 28/11/2011 19.49.24
Hi,
because I found QGIS to be terribly slow with larger vectors, I
started to work a bit on rendering optimization. I am still at the
beginning of the work, here are some first findings.
[CUT ALL]
These benchmarks are
Hi,
because I found QGIS to be terribly slow with larger vectors, I
started to work a bit on rendering optimization. I am still at the
beginning of the work, here are some first findings.
The first step was to get some reliable benchmarks, for which I wrote
a new CLI application, mostly derived
31 matches
Mail list logo
