Hello all,
I tried to fiddle with the runcode.py script but it's beyond my reach unless I take some python tutorials.
For the moment I have settled running jobs locally, which works fine. It would have been great to harness the capabilities of the Job Scheduler and Microsoft HPC Pack to distribute workload among the workstations of the cluster.
The good thing is that it is possible to compile opentelemac with
mingw-64 and
MS-MPI library. Development of MPICH2 has stopped for windows and Microsoft with MS-MPI has taken over. Having said that, I wouldn't recommend users to switch to MS-MPI unless they plan to set up a HPC Cluster by installing MS HPC Pack (which is free
but requires a Server product to act as a headnode). Even then, it seems that telemac needs some adapting if it is to submit jobs seamlessly.
I would be happy to help, to the extends of my capabilities, with the development/testing of telemac to adapt to Microsoft HPC Pack.
Best Regards,
Costas