How to run a multi-core job on the grid

Summary

This article outlines how to submit a multi-core job on the Grid. It involves copying a sample script (multicore.sh), editing it to include your email, and submitting it with sbatch.

Body

Follow the steps below to run a multi-core job on the Grid.

Step 1

Log on to the Grid.

Step 2

From your home directory, copy the multicore job script to your home directory by typing cp /wsu/el7/scripts/tutorial/multicore.sh.

Step 3

View the contents of your home directory by typing ls.

Step 4

Edit the script using vim by typing vim multicore.sh.

Step 5
  • Press i to insert and use the up and down arrow keys to scroll through it.
  • Be sure to change the email address to your own.
  • When finished editing type wq and Enter to save and quit.
Step 6
  • Run the multi-core job script by typing sbatch multicore.sh.
  • In Submitted batch job 2626008, 2626008 is the job ID.
Step 7

Check to see that your job is running by typing qme.

Step 8
  • You can log in to the node that your job is running on by typing ssh $HOSTNAME (where $HOSTNAME is the name of the node your job is running on).
  • In this example, ssh can be done to node wsu196 by typing ssh wsu196.
  • You'll be prompted to enter your password. Once you're on the compute node, you can check to see how many resources your job is using by typing top -u $USER (where $USER is your AccessID)
  • Hit q to quit top and type exit to leave the compute node.
Step 9
  • When your job is finished, you will have an error and output file in your home directory.
  • The files are formatted as output_<jobid>.out and error_<jobid>.err.
  • Check by typing ls.
  • You can check the contents by using the command cat.
  • For this example, it would be cat output_2626008.out
  • The outputs of the script are the working directory, the hostname, the date, and the number of cores the job is running on.

Details

Details

Article ID: 20220
Created
Wed 7/16/25 11:40 AM
Modified
Wed 11/19/25 2:18 PM