Hi Robert,

On Thu, May 22, 2008 at 1:44 PM, Robert Osfield <[EMAIL PROTECTED]>
wrote:

>
> I am also particularly interested in feedback from yourself on the
> performance of the new DatabasePager
> and curl plugin on your own databases.  These changes should mean that
> you should get the performance
> you want from standard OSG components rather than having to utilise
> custom curl plugin.
>

I'll test as soon as I can, but currently we are in a big refactoring phase
of our work which means that the http part is currently broken. :) It will
allow me to directly use the standard OSG components and test them instead
of making our own plugin work again. I'll let you know when it will be done.


>
> Out of curiosity, how many threads did you allocate to your curl
> plugin to do the http reading?  Did you do
> any performance benchmarks on single vs multiple http threads?  The
> new curl plugin can handle multiple
> threads so we can test multiple http threads with the new set up -
> I've tested the functionality and it works, but
> haven't done any benchmarking of this configuration yet.
>
>
We've made lots of performances tests from 1 thread with a curl multi
handler to 20 threads with single handlers. Let's say that the version with
only one multi handler thread is finally the better one as having lots of
threads brought us lot of other problems.

One thing we never test is multiple http threads with each multi handlers, I
don't know if it can really boost the performances.

But from our experience, with http request, the most important to have a
fluid experience is to request the right nodes at the right moment, and
really try to minimize the useless calls. I even though to a system to be
able to request nodes even if there parents nodes are not still loaded (I am
not sure but I think Google Earth use this system).

-- 
Serge Lages
http://www.tharsis-software.com
_______________________________________________
osg-users mailing list
[email protected]
http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org

Reply via email to