[mapserver-users] is mapserver on windows slower ???
Tyler Mitchell
TMitchell at lignum.com
Wed Jun 26 11:19:42 PDT 2002
Hi Gregor, from Canada
>From my experience with other applications I have rarely seen (never?) a
service use more than 50% max on each cpu at a time. If there were
parallel requests going on I assume there would be some performance
improvement with the dual chip. But don't applications have to be written
to take advantage of this? Is this called multi-threading?
Tyler
Gregor Friedrich
<gregor at hdg.ufz.de> To: mapserver-users at lists.gis.umn.edu
Sent by: cc:
owner-mapserver-users at lists. Fax to:
gis.umn.edu Subject: [mapserver-users] is mapserver on windows slower
???
06/26/2002 09:11 AM
Hi there,
If am testing the same map services with identical data sources (shapefiles
with *.qix index) one serving on Linux (SUSE 7.3, Apache 1.3.19... self
compiled binary ) the other one on Win2k Server with IIS 5.0 etc,
downloaded binary ) , the Win machine have 2 Pentium II 300 Mhz each, the
Linux box just one Pentinum II 300 Mhz . I had expected that the WIn2k box
is 2 times faster around than the Linux box , but the win box needs more
time to build the maps. The service on Linux needs 20 seconds (64000
polygons !) to start up , the windows service 30 s . On Linux mapserv
processes consume > 90 % of processor power, on windows only 50 - 54 % , it
seems that the parallelizing is n't so good , is n't it ?
greetings from germany
Gregor
More information about the MapServer-users
mailing list