Hello all,
So far, when I want to run a parallel computation in a different node than the local one, I must hardcode it in the 'mpi_cmdexec' entry in the configuration.
According to 'runcode.py' one can define a hosts list using the '--hosts' argument. However, there is no means for the hosts list defined with the '--hosts' argument to appear in the 'mpi_cmdexec' command reconstruction like it is done for the HPC job submission (at least to my understanding).
Is this kept that way for some reason? Should I modify the execution scripts to obtain this functionality or is this something that can be added to the trunk?
Best Regards,
Costas