Depending on what you're doing, any bulk requests should be to a
pre-seeded MapCache and not the WMS server per se. In a production
environment a modestly detailed raster, it's easy to kill a WMS server
of any size.
You're going to have to test a range of requests to get a usable setup.
You
As Oskar wrote, there is no single answer to this question. It depends
on lots of factors.
To give you an idea, you may be interested in checking out the FOSS4G
WMS Benchmarks [1], a friendly competition that took place for a few
years at FOSS4G comparing MapServer and other contenders under
It is impossible to answer on without knowing what kind of for
wms/wcs/wfs you are serving with mapserver
For example a simple raster wms will be able to do many request per second.
On the other side serve out osm from a postgis database will in some
heavy scale not do more then 1 tile every
Hello
I'm looking for a benchmark advice.
If I have a machine running on 6 CPUs and 25gb RAM what would be the value of
the number of requests per second from which I should start scaling my VMs with
Mapserver?
Is there any way of testing or investingating? I'm using Locust and JMeter but
