@@ -12,7 +12,7 @@ This example uses local storage as we are not dealing with large amounts of data
## Running Jupyter Notebook
To begin, SSH to the login node. Instructions on how to connect are provided in the welcome email.
Jupyter Notebook is started on the cluster like any other workload, by submitting a job through Slurm.
As the compute nodes (where workloads run on the cluster) are not directly reachable from the campus network, we need to set up a chain of SSH port forwards to access Jupyter Notebook instances. The job script will:
1. Start a Jupyter notebook server on an available port on the compute node
...
...
@@ -21,7 +21,7 @@ As the compute nodes (where workloads run on the cluster) are not directly reach
- The login node
- Finally reaching your compute node
Save the following script as `jupyter.sbatch`:
On the login node, save the following script as `jupyter.sbatch`:
