You have not really explained your issues in a detail sufficient to
give any practical advice. As always, analyze the problem for areas
of parallelism. There is no built-in mechanism for MPI in Mapserver,
but if you wanted to really bust up the code, the obvious place of
inefficiency is the fact that all layers are drawn serially.
A hack to get around that is to bust up your mapping file, into one
layer per map file. Then give each layer to a different machine.
Then write a new "master" map file that reads from all the children
as WMS layers. Put that on yet another machine. Because Mapserver
sends out the WMS requests in parallel you can get all the layers
rendered simultaneously on your farm, and then assembled at the end
on the master.
Paul
On Jan 19, 2006, at 8:10 AM, Biz King wrote:
Hi All.
Is anyone aware of anywhere (or better still, has experience of)
running Mapserver via an MPI/Grid interface or as a cluster?
We're trying to develop a high-performance mapserver that can cope
with the load we're going to be throwing at it! Currently it takes
298 seconds (on a Mac OSX Server, 3.5 Gb Ram, dual 2Ghz processors)
to do what we need done on under 60 seconds! There's not much we
can do to cut down the load as we're creating a whole series of
nodes on a layer via a database and we're then creating the imagery
based on these items and outputting them to graphics formats in
varying sizes.
The results get fed to users on demand without the delays
associated with 'on the fly' image creation.
Any help will be welcomed!
cheers
Biz
