mpiexec - PMI job lauch (mpich2)

[Johnny Devaprasad]

Hi,
When i try to run an mpich2 job with the following line
   mpiexec --comm=pmi -n 2  ./test  (because my default mpiexec is built
with --with-default-comm=mpich-p4)
      (I have also tried the above after rebuilding mpiexec with
--with-default-comm=mpi)

mpiexec seems to spawn two different processes which dont belong to the same
process group. The output of my program is
   Hello world from process 0 of 1 from node002
   Hello world from process 0 of 1 from node001

My test program is as follows:
======================
#include <stdio.h>
#include <mpi.h>

int main (int argc, char **argv){
  int rank, size, name_length;
  char processor_name[MPI_MAX_PROCESSOR_NAME];
  MPI_Init (&argc, &argv);/* starts MPI */
  MPI_Comm_rank (MPI_COMM_WORLD, &rank);/* get current process id */
  MPI_Comm_size (MPI_COMM_WORLD, &size);/* get number of processes */
  MPI_Get_processor_name(processor_name, &name_length);
  printf( "Hello world from process %d of %d from %s\n", rank, size,
processor_name );
  MPI_Finalize();
  return 0;
}
===========================

But, when i run it outside the queueing system, with the 'mpiexec' provided
by mpich2, it is working fine
  mpiexec -n 2 -machinefile machines ./test

The output is as follows:
   Hello world from process 1 of 2 from node002
   Hello world from process 0 of 2 from node001


Other details:
==========
mpiexec version - 0.82
torque - 2.1.8
mpich2 - 1.0.5p4 (compiled with gcc)
cluster - Scientific Linux 5 based VMware cluster.

Any solutions would be greatly appreciated.

Regards,
Johnny
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://email.osc.edu/pipermail/mpiexec/attachments/20070913/94b95d31/attachment.htm