From: Josh Guffin Newsgroups: comp.parallel.mpi Subject: Strange occurance Date: Wed, 14 Jul 1999 08:52:35 -0500 Organization: Purdue University Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Xref: ukc comp.parallel.mpi:5322 (I'm using MPICH over a network of RH Linux 6.0 boxen) I am attempting to run a simple program which writes the ID # of a processor to a file called data.out on that machine. Two of my machines are down, so i'm using the -machinefile option to specify the machines that are still up... For some reason, the binary is apparently not being executed on the last machine in the table. Here is what my machinefile looks like: meco01 meco02 meco03 meco04 meco05 meco07 meco09 I'm using rdist to send the source to each machine and compile it there. The code is: program main include "mpif.h" integer myid call MPI_INIT(ierr) call MPI_COMM_RANK(MPI_COMM_WORLD, * myid,ierr) open(unit=5,file="data.out") write(5,101)myid 101 format(1x," my id number is = ",i10) call MPI_FINALIZE(ierr) stop end It is compiled with mpif77 on each machine. Once there i use the command: mpirun -np 7 -machinefile machines a.out The last processor in the list never writes data, and is apparently not even being run. Any ideas as to why this is? Thanks, Josh =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= = Josh Guffin guffin@purdue.edu = = Purdue U. HEP - TASK E expert.cc.purdue.edu/~guffin = = #include = =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=