you are using mpich version of Firefly and it should be run in
parallel by mpirun exactly as expected by mpich. It seems your
mpirun is of different MPI implementation. Ask your system administrator
how to run mpich binaries in parallel on your cluster.
Alternatively, you can manually start mpich-linked Firefly in
parallel without any mpirun but rather using proper procgrp file
and passing -p4pg full_name_of_procrgp_file mpich option at the
end of the command line.
Hope this helps.
Note, there are multiple prior discussions of various mpich-related
problems on this forum. They may be of interest to you.
On Sat Sep 4 '10 3:11pm, Olga wrote
>I work with Firefly on cluster, which has only 64-bit MPI and Cleo manager. I use Firefly version 7.1.G, Linux MPICH (using rsh as remote shell by default), fully statically linked.
>My input string:
>mpirun -np 8 /home/kremov/ff/firefly -r -f -p -stdext -i /home/kremov/ff/1/bench01.inp -o /home/kremov/ff/1/bench01.out -ex /home/kremov/ff -t /home/kremov/tmp
>My task takes up 8 cores, but in out-file I see:
>1) *ERR* IO-09 system file error - No such file or directory
>2) PARALLEL VERSION (MPICH) RUNNING IN SERIAL MODE USING SINGLE PROCESS
>3) Warning: running without fastdiag runtime extension!
>Why does it happens? Is the cause of this 64-bit MPI?