I encountered the following error when running rmcdhf_mem_mpi on the HPC cluster I work on:
Default settings? (y/n)
Error termination. Backtrace:
At line 33 of file /home/USER/code/grasp/src/appl/rmcdhf90_mem_mpi/systemmem.f90
Fortran runtime error: Integer overflow while reading item 1
Could this mean that the amount of system memory SYSTEMMem() reads from /proc/meminfo is too large to fit into an integer? Perhaps this could be fixed by changing the kind of SUMMem? On our cluster, I've observed this for jobs that assigned 1024 GB or more to a node - 512 GB worked fine.
I encountered the following error when running
rmcdhf_mem_mpion the HPC cluster I work on:Could this mean that the amount of system memory
SYSTEMMem()reads from/proc/meminfois too large to fit into an integer? Perhaps this could be fixed by changing the kind ofSUMMem? On our cluster, I've observed this for jobs that assigned 1024 GB or more to a node - 512 GB worked fine.