Not understanding how to enable multiproc dump files

Hi All,

I am working on parallel file systems and want to try out some different science applications on them. I took the LAMMPS version from the CORAL site (7-12-2016) and am trying some of the examples (i.e., in.crack).

Basically what I want is, for each MPI process to output it’s atoms array (and one file per timestep), but I am not sure how to modify this line (from examples/in.crack):

dump 1 all atom 500 dump.crack

I have read the web page and some of the header files, but it’s not clear to me what I need to do. Any help is appreciated.

Regards,
Josh

Hi All,

I am working on parallel file systems and want to try out some different
science applications on them. I took the LAMMPS version from the CORAL site
(7-12-2016) and am trying some of the examples (i.e., in.crack).

Basically what I want is, for each MPI process to output it's atoms array
(and one file per timestep), but I am not sure how to modify this line
(from examples/in.crack):

dump 1 all atom 500 dump.crack

I have read the web page and some of the header files, but it's not clear
to me what I need to do. Any help is appreciated.

​pretty straightforward:​

​if you include a character ​in the file name, then you will get one file per processor and the will be replaced by the MPI rank of the process
writing it. if you insert a * character in the file name, it will be
replaced by the current time step number.

axel.

That worked, exactly how I hoped. Thanks.