Ginsburg: Getting Started

Getting Access

Access to the cluster is subject to formal approval by selected members of the participating research groups. See the HPC Service Webpage for more information on access options.

Introductory Linux Training and Resources

If you're not familiar with basic Linux commands and usage, or if you need a refresher on these topics, please refer to the following resources from our workshop series:

• Intro to Linux - Slides and Video Recording A general intro to the Linux command line. Not about Ginsburg.
• Intro to Shell Scripting - Slides and Video Recording A general intro to the Linux command line. Not about Ginsburg.
• Intro to High Performance Computing - Slides and Video Recording Computing concepts specifically for clusters, such as Ginsburg

For a list of recorded trainings and upcoming research computing workshops and events, please see:

https://www.cuit.columbia.edu/rcs/training

Logging In

You will need to use SSH (Secure Shell) in order to access the cluster.  Windows users can use PuTTY or Cygwin. MacOS users can use the built-in Terminal application.

Users log in to the cluster's submit node, located at ginsburg.rcs.columbia.edu or use the shorter form burg.rcs.columbia.edu.  If logging in from a command line, type:

$ ssh <UNI>@ginsburg.rcs.columbia.edu



OR


$ ssh <UNI>@burg.rcs.columbia.edu

where <UNI> is your Columbia UNI. Please make sure not to include the angle brackets ('<' and' >') in your command; they only represent UNI as a variable entity.

Once prompted,  you need to provide your usual Columbia password.

Submit Account

You must specify your account whenever you submit a job to the cluster. You can use the following table to identify the account name to use.

Your First Cluster Job

When you first login to Ginsburg, you are on a login node. Login nodes are not places where users should do actual work aside from simple tasks like editing a file or creating new folders.

Instead, it is important to move from the initial login node to a compute node before doing most work. Example:

srun --pty -t 0-2:00 -A <ACCOUNT> /bin/bash

Now you have moved from the login node to one of the compute nodes on the cluster.  The simple tasks mentioned above can also be done here, but from here is where it is especially important to submit scripts for processing.

If the HPC group notices jobs being run on a login node, such jobs will be terminated and the user notified.

Submit Scripts

This script will print "Hello World", sleep for 10 seconds, and then print the time and date. The output will be written to a file in your current directory.

In order for this example to work you need to replace ACCOUNT with your group account name. If you don't know your account name the table in the previous section might help.

#!/bin/sh
#
# Simple "Hello World" submit script for Slurm.
#
# Replace ACCOUNT with your account name before submitting.
#
#SBATCH --account=ACCOUNT        # Replace ACCOUNT with your group account name 
#SBATCH --job-name=HelloWorld    # The job name
#SBATCH -N 1                     # The number of nodes to request
#SBATCH -c 1                     # The number of cpu cores to use (up to 32 cores per server)
#SBATCH --time=0-0:30            # The time the job will take to run in D-HH:MM
#SBATCH --mem-per-cpu=5G         # The memory the job will use per cpu core

echo "Hello World"
sleep 10
date

# End of script

Job Submission

If this script is saved as helloworld.sh you can submit it to the cluster with:

$ sbatch helloworld.sh

This job will create one output file name slurm-####.out, where the #'s will be replaced by the job ID assigned by Slurm. If all goes well the file will contain the words "Hello World" and the current date and time.

See the Slurm Quick Start Guide for a more in-depth introduction on using the Slurm scheduler.