search

qusers / qligfep

Other
48 stars 23 forks source link

QligFEP inputfiles/runscript bug #11

Open bananatana opened 1 year ago

bananatana commented 1 year ago

QligFEP automatically generates runCLUSTER.sh within inputfiles. The problem is that generated run.sh always contains:

time mpirun -np 32 $qdyn eq1.inp > eq1.log (for example)

which will run for example on CSB, but not on Tetralith where mpirun should be replaced with mpprun

it can be replaced manually with sed -i 's/mpi/mpp/g' run.sh but it is not really convenient if there is a lot of molecules.

jesperswillem commented 1 year ago

We could add it to the settings.py file?

bananatana commented 1 year ago

We could. I will try, if u have any idea of solution on mind please let me know (:

uto, 21. ožu 2023. u 14:17 jesperswillem @.***> napisao je:

We could add it to the settings.py file?

— Reply to this email directly, view it on GitHub https://github.com/qusers/qligfep/issues/11#issuecomment-1477823036, or unsubscribe https://github.com/notifications/unsubscribe-auth/A6KUY4KNYLPCHJOX7GEPGN3W5GTADANCNFSM6AAAAAAWCNQWUE . You are receiving this because you authored the thread.Message ID: @.***>

-- Tana Tandarić, Ph.D. Post-doctoral fellow at Department of Cell and Molecular Biology, Computational Biology and Bioinformatics Uppsala University Uppsala, Sweden

Tana Tandarić, Ph.D. Post-doctoral fellow at Department of Organic Chemistry and Biochemistry Ruđer Bošković Institute Zagreb, Croatia

afloresep commented 1 year ago

I have solved this in my fork by adding a condition to QligFEP, so that when TETRA is used as an argument, the outline becomes "mpp" instead of "mpi".

jesperswillem commented 1 year ago

perfect, that was precisely the first fix I had in mind. Perhaps we need better management of cluster setup later on though (config files maybe?) Would be good to discuss a direction there.

afloresep commented 1 year ago

Any way I can help let me know