[gmx-users] Perplexed about MPI

[dr.horsfield at physics.org]

Hi,

I have begun to despair of MPI! I have LAM-MPI 6.5.7 installed. It passes
the test suite perfectly, and the Pallas MPI benchmarks run fine. I have
gromacs 3.1.4 installed, and I have compiled it for MPI. A particular
small job runs fine on 2 processors (either on the same box, or on two 
boxes), but falls over immediately on 3 or 4, with the error:

MPI process rank 0 (n0, p3603) caught a SIGSEGV.
-----------------------------------------------------------------------------

One of the processes started by mpirun has exited with a nonzero exit
code.  This typically indicates that the process finished in error.
If your process did not finish in error, be sure to include a "return
0" or "exit(0)" in your C code before exiting the application.

PID 3603 failed on node n0 with exit status 1.
-----------------------------------------------------------------------------

Could the problem be that the number of atoms on each node is not the same 
for 3 or 4 processors? Any suggestions?

Cheers,

Andrew