Hello
I'm a new user in HPC. The model I'm working on was written on my laptop and ran appropriately there. On my laptop, I compiled the "make auto" file to generate lmp_auto. But on the cluster, the file compiled was "make mpi". The model I am trying to run has two parts; the first one is the generation of particles, where I create a restart file; this stage works appropriately. Then, I run another script where I import the restart file and run the loading stage. There appears the problem because, although I run the script with the same number of processors and the same dimensional distribution of them on both scripts, the number of particles read on the second stage is not the same. For example, I generated 26000 particles in the early stage; on the second one, there are only 4000. The scripts I am running are the same I tested on my laptop; the only difference is the number of processors and its distribution. I have tried to run the same version I generated and successfully run on my laptop, with the same processors' distribution, but the problem persists. And I also tried to run the script just with one processor, and I get the same problem. Could someone help me, please?
I don't know if the problem could be related to this warning:
WARNING: OMP_NUM_THREADS environment is not set. (../comm.cpp:146)
using 1 OpenMP thread(s) per MPI task
Thanks
Kind Regards
Miguel