menu

Running a batch job

GaussView is installed and can be used on the following shared servers

  hostname: kagayaki

How to run a Gaussian16 batch job

How to run a Gaussian16 batch job:

1. Login to the server.

2. Change the current directory to job data dir.

%cd /home/<username>/XXXXX

input file example:

#p bp86/6-31G** Opt=(ModRedundant)

C2H4

0 1
C1             0.709378            0.005672           -1.977566
C2             1.675673            0.033777           -1.062471
H1             0.311648           -0.935823           -2.365568
H2             0.275384            0.922568           -2.383892
H3             2.073809            0.975215           -0.674624
H4             2.109710           -0.883057           -0.655940

3. Copy the sample batch script and your input file to your working directory.
A sample batch script is below.

Sample of Batch Script:

/Samples/Gaussian/

% mkdir Gaussian
% cp /Samples/Gaussian/g16/*  /home/<username>/Gaussian
% cd Gaussian

6. Confirm the setting of the batch script (PBS_g16.sh).

#!/bin/sh
#PBS -j oe
#PBS -l select=1:ncpus=16:mpiprocs=16   :Use 16Cores ## (DON't change "select=1" parameter)
#PBS -q SINGLE            : Submit job to SINGLE queue.

module load g16/c.01

cd $PBS_O_WORKDIR

echo "-P-16"  > Default.Route   : for 16Core job, the number should be 16

g16 < gaussian.com > gaussian.log   : correct the filename "gaussian.com" to your inputfile name.

rm -f ./Default.Route

7. Submit the job

% qsub PBS_g16.sh

 

For other information about running batch jobs, please refer to the manual for parallel computing.

8. The result file will be shown after the job finishes.
In the above batch script , the output result file name was called "gaussian.log".