Local Options 

-n <# of processes> or -np <# of processes> 

Use this option to set the number of MPI processes to run with the current argument set.

-env <ENVVAR> <value> 

Use this option to set the <ENVVAR> environment variable to the specified <value> for all MPI processes in the current arg-set.

-envall 

Use this option to propagate all environment variables in the current arg-set.

See the I_MPI_HYDRA_ENV environment variable for more details.

-envnone 

Use this option to suppress propagation of any environment variables to the MPI processes in the current arg-set. 

-envlist <list of env var names> 

Use this option to pass a list of environment variables with their current values. <list of env var names> is a comma separated list of environment variables to be sent to the MPI processes. 

-host <nodename> 

Use this option to specify a particular <nodename> on which the MPI processes are to be run. For example, the following command executes a.out on hosts host1 and host2:

$ mpiexec.hydra -n 2 -host host1 ./a.out : -n 2 -host host2 ./a.out

-path <directory> 

Use this option to specify the path to the <executable> file to be run in the current arg-set.

-wdir <directory> 

Use this option to specify the working directory in which the <executable> file runs in the current arg-set.

-umask <umask> 

Use this option to perform the umask <umask> command for the remote <executable> file.

-hostos <host OS>

Use this option to specify an operating system installed on a particular host. MPI processes are launched on each host in accordance with this option specified. The default value is linux.

Arguments

<arg> 

String parameter

linux

The host with Linux* OS installed. This is the default value

windows

The host with Windows* OS installed.

Note:

The option is used in conjunction with –host option. For instance, the following command runs the executable a.exe on host1 and b.out on host2:

$ mpiexec.hydra -n 1 -host host1 –hostos windows a.exe : -n 1 –host host2 \ –hostos linux ./a.out