On Sep 15, 2010, at 5:12 PM, Artem Pavlenko wrote: > Hi Richard, > > The 'postgis' graph is showing clearly that out of all WMS > participating in that test, only Mapnik (paleoserver) scales beyond 64 > concurrent requests, which is a very good result, indeed. > > Regards, > Artem
One excellent thing about the benchmarking exercise is that Michael Smith at the US Army Corps of Engineers, who kindly donated the hardware, has committed to keep the servers on and running. Huge thanks must go to Michael and his org for this. Most notably it will give each team of server representatives that participated a chance to more fully digest results and come up with better types of tests - that are more fair and representative. Also it will allow me in the coming weeks to test have far Mapnik can keep on scaling beyond 100+ concurrent client requests. There has got to be an important drop off point where the server becomes maxed out, but as Artem points out we've not hit that yet in this simple test. - Dane _______________________________________________ Mapnik-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/mapnik-users
