RE : [Benchmarking] Consensus

[Julien-Samuel Lacroix]

Hi,

Thank you very much for this.

I'll leave others comment on that and possibly integrate it in the slides.

I think this reflects the facts we discuss yesterday. If there's other 
things to say I don't know since I was not in the project enough so I'll 
let others comments.

Thanks
Julien

On 10-09-08 04:23 AM, Luc Donea wrote:
> Hi Julien,
>
> Thanks for putting all this together.
>
> I think we should add slides to explain what we discussed yesterday, in
> the common (team agnostic) part of the presentation.
>
> I would propose the following content, please feel free to comment/edit:
>
> * The initial bottleneck in the benchmark tests was the disk read access
> * Some teams managed to turn this 'disk-bound' scenario into a
> 'disk-unbound' scenario, allowing servers to fully run in Memory with no
> disc access, putting all load on the CPU.
> * Results observed in this 'disk-unbound' scenario were 5 to 10 times
> better than in the 'disk-bound' scenario.
> * This was possible thanks to the conjunction of the following:
> ** the Red Hat Enterprise Linux sever is efficiently caching data blocks
> at the OS level
> ** the 2000 WMS requests are the exact same between all runs
> ** the Server RAM memory is of 8GB
> * Unfortunately, the Windows server did not behave the same way as RHEL.
> * The Windows server was down in the last benchmark days, preventing
> some teams to try to reproduce the disk unbound scenario.
> * As apples cannot be compared to oranges, all participants did agree
> that teams were not mandatory to publish their benchmark results in the
> final presentation.
>
> I have put the proposal into the attached slides.
>
> Thanks,
>
> Luc
>
>
> -------- Message d'origine--------
> De: benchmarking-bounces at lists.osgeo.org de la part de Julien-Samuel Lacroix
> Date: mar. 07/09/2010 20:29
> À: Performance testing of OSGeo and other web service engines.
> Objet : [Benchmarking] Consensus
>
> Mike (or anyone really), please correct me if I'm wrong.
>
> Hi,
>
> A quick reminder to every teams. You have to provide around 3 or 4
> slides for the presentation tomorrow before noon. Please integrate them
> in the Google Doc Presentation.
>
> Some idea of what you can talk about:
> Why it was so great?
> Difficulties?
> Any enhancement you may have made?
> Any tips and tricks for your users?
>
> You don't have to includes results in those slides. It's up to you.
> There will be a comparison of the results in the presentation at some
> point. Please email on the list if you don't want your results to be
> included in the charts that will be presented (if you didn't explicitly
> tell so during the meeting).
>
> Here's the slide list (not necessarily in order):
> - History
> - Fourth exercise
> - First time with windows
> - First time with something other than MapServer and GeoServer
>
> - Summary
>
> - 4 slides by Ivan describing the dataset (with the correction on the
> GeoTiff slide)
>
> - Rules of engagement
>
> - Everyone prepare 4 slides.
>
> - Results
>
> - What's next
> - Setup a process to discuss next year exercise
> - Probably change the data to local data (Coloroda) next year
> - etc
>
> - Clear link to the wiki that there's more info there on some exercices
> and on team's results.
>
>
> I hope that helps.
>
> Julien
> --
> Julien-Samuel Lacroix
> Mapgears
> http://www.mapgears.com/
> _______________________________________________
> Benchmarking mailing list
> Benchmarking at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/benchmarking
>
>
>
>
> _______________________________________________
> Benchmarking mailing list
> Benchmarking at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/benchmarking

-- 
Julien-Samuel Lacroix
Mapgears
http://www.mapgears.com/