Dear ALL,
I encountered a strange problem.
My input file works well on Windows LAMMPS.
When I use linux MPI-version, it also works well for “-np 4”.
However, when I run using “-np 8”, it fails with below error messages, after running 1000 MD steps.
rank 2 in job 4 server104_45264 caused collective abort of all ranks
exit status of rank 2: killed by signal 11
Did anyone encounter the similar problem?
Thanks and best regards,
hengan
2010/11/19 Hengan Wu <[email protected]...>:
Dear ALL,
I encountered a strange problem.
My input file works well on Windows LAMMPS.When I use linux MPI-version, it also works well for "-np 4".
However, when I run using "-np 8", it fails with below error messages, after
running 1000 MD steps.rank 2 in job 4 server104_45264 caused collective abort of all ranks
exit status of rank 2: killed by signal 11
signal 11 means: segmentation fault.
this is a very non-descript error.
in fact, from the (too) few lines you quote, you seem to be
using MPICH and in that case a segmentation fault
_always_ happens when the simulation aborts from
some reason.
now to find out the reason, you either have to contact
somebody with psychic reading ability or provide more
information about the job you are running, starting with
the lines of output leading up to this error.
axel.