Calculations stopping without error message

[Matthew Graneri]

Hi there,

I'm a very new NWChem user and I have come across something that might either be a bug with the program itself or the installation on our computer cluster (installed with EasyBuild), I'm not sure which.

Basically, while I have been able to run two calculations to completion, two others I have tried have died within seconds of being started. Surprisingly, this includes one of the calculations included in the NWChem QA test set (the h2o-b3lyp-disp.nw input). I've attached all the output I get below—as you can see, no error message. There are also no error messages in the cluster's Slurm output—as far as Slurm is concerned, the calculation completed successfully. It also makes no difference whether you run it with mpirun or not.

I was hoping someone would be able to confirm on their version of NWChem whether they are having the same issue? Or, if you see anything that looks wrong in the input file that will not work with modern NWChem (because it was written a long time ago), would someone be able to let me know what should be adjusted?

[Edoardo Aprà]

This  clearly shows that the NWChem Easybuild installation on your system is not working.
I have just tried to run NWChem  on Almalinux 8.4 using the EPEL rpm (NWChem 7.0.2) and the docker  and singularity/apptainer images (for 7.2.3). In all this cases NWChem ran to completion