Q/A

Q1.Dear All,

I am trying to run my job but Penzias doesn't accept it. I need to try a 16 nodes memory. Let me know, please what is my mistake. Below you have my job scripts:

1. !/bin/bash
2. SLURM -q production
3. SLURM -N test_nwchem
4. SLURM -l select=2:ncpus=16:mem=23040mb
5. SLURM -l place=free
6. SLURM -V

cd $SLURM_O_WORKDIR

A1. There are 2 mistakes here. From SLURM point of view there is no nodes with 16 cores per node. The SLURM "thinks" that all nodes have 8 cores. Second the maximum available memory per core is 3686 mb.

Q2. I am trying to submit a job but que production_qdr does not exist.

A2. On penzias there is no que name production_qdr. On PENZIAS the main que is called production. This was marked in text and in example script on NWChem wiki.