[Paradiseo-help] [paradiseo-Bugs][8809] Need further details regarding the PEO implementation in Tutorial

paradiseo-bugs at gforge.inria.fr paradiseo-bugs at gforge.inria.fr
Mer 21 Oct 14:47:31 CEST 2009


Bugs item #8809, was opened at 2009-10-21 14:47
Status: Open
Priority: 3
Submitted By: Caner Candan (candan_c)
Assigned to: Nobody (None)
Summary: Need further details regarding the PEO implementation in Tutorial 
Category: None
Group: 1.2.*
Resolution: None


Initial Comment:
I have applied all what is written at the tutorial pages of PEO in using 4 machines with 4 CPUs by machine but I have got a strange behavior:

As I have defined Node1 as the runner (keeping in mind nodes starts by 0) in my schema.xml file the only processor used is in this machine and the others ones stay in standby.

My question is: Is it normal ?

YES -> then could you explain in a few lines why and how MPI works with the framework. I think it's going to be better if you add more details about this topic in the documentation.

NO -> let's say why ?

Here's the source of the file schema.xml:

<?xml version="1.0"?>

<schema>
     <group scheduler="0">
          <node name="0" num_workers="0"></node>
          <node name="1" num_workers="0">
                <runner>1</runner>
          </node>
          <node name="2" num_workers="1"></node>
          <node name="3" num_workers="1"></node>
      </group>
</schema> 

The hosts file used by `mpdboot` command:

mach0:4
mach1:4
mach2:4
mach3:4

command syntax: `mpdboot -n 4 -machinefile hosts` to start mpds

BR,
Caner Candan


----------------------------------------------------------------------

You can respond by visiting: 
http://gforge.inria.fr/tracker/?func=detail&atid=663&aid=8809&group_id=145




Plus d'informations sur la liste de diffusion Paradiseo-help