[MapProxy] concurrent seed workers on windows

[Brunner Tobias (GEO)]

At the moment, we are testing mapproxy (in combination with QGIS Server) on both, windows and linux.

Since QGIS Server is faster in providing the WMS images than Mapproxy is in saving the tiles, we would like to use multiple concurrent seed workers. When using mapproxy on linux this is no problem: we can drive the CPU usage of the QGIS Server (both windows and linux) up to 100% when using around 24 concurrent seed workers.

When using mapproxy on windows though, mapproxy can only drive the CPU usage of the QGIS Server to an average of 20% - with spikes reaching around 60%, followed by inactivity of about 1-2 seconds. On the mapproxy-machine, there is a single python.exe process running which uses a maximum of about 26% of the CPU and never more than 700MB of ram.

I suspect this may have something to do with the python-multiprocessing on windows...?
Is this a known issue on windows and are there any workarounds?


Systems:
Mapproxy-Windows-VM: 8 cores, 12 GB of ram, windows server 2012 r2, python 2.7 (osgeo4w)
Mapproxy-linux-VM: 8 cores, 12 gb of ram, redhat

I've also tested this on a windows 2008 r2 machine with 4 cores, 4 gb of ram and python 64bit ... resulting in the same behavior with the cpu load around 40 percent.

Thanks for your help!

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/mapproxy/attachments/20151222/2560134d/attachment.html>