I’m trying to run CalculiX (PrePoMax) on an HPC server and I’m getting an error – it isn’t even getting to iterations. But when I run on my local laptop it runs fine and produces results. Have tried the PrePoMax community and they recon it’s best to ask here.
Maybe the HPC server does not have the requested PARDISO related libraries. As next step I would recommend to run the inp file without a dedicated solver by replacing:
*Heat transfer, Solver=Pardiso, Steady state
through:
*Heat transfer, Steady state
which would then use the default solver on the selected computer. Thank you!
Tried that and it’s still failing. I would’ve posted the .log but it’s not letting me
I also tried changing the solver to Default in PrePoMax and it is also failing
Did you try running any other model / .inp file (e.g. one from the test examples)?
From your log file it seems to end while trying to write this info message:
*INFO in gentiedmpc:
failed nodes are stored in file
CutStone1_BC_NoIns_WarnNodeMissTiedContact.nam
This file can be loaded into
an active cgx-session by typing
read CutStone1_BC_NoIns_WarnNodeMissTiedContact.nam inp
Is the stderr also directed to this log file or do you have any further output?
I’ve used > FileName.log to output the log
Can you run > stdout.log 2> stderr.log and check if something ends up in the latter?
What does the exit code say?
echo $?
stderr.log is empty. In the stdout.log, here are the last few lines:
The strange thing is that whenever I run the simulation, it goes into a CalculiX cmd window rather than doing it within PrePoMax (as is the case with my local laptop)
I have no experience with compiling and running CalculiX on Windows, but in general running the executable on a different hardware / software environment (other than where it was compiled) can be a problem.