Use Jupyter Notebook on any server

This guide explains how to run a Jupyter Notebook on a department server and access it from your browser. The steps differ depending on whether the server has a public IP address or not.

Servers without public IP addresses (Turing, Kepler, Ada, Ampere, Neumann)

These servers require SSH tunneling through a public-facing server. Turing is used as an example here — replace it with the server you are using.

1. SSH into Tesla (a public-facing server).
2. From Tesla, SSH into Turing.
3. Create a conda environment on Turing and activate it.
   Read instructions
4. Install Jupyter: conda install jupyter
5. Start the notebook: jupyter notebook --no-browser --port=8889 --ip=0.0.0.0
6. You will see a token in the output. Copy it.
7. On your local machine, open WSL (Windows) or Terminal (Linux/Mac) and set up an SSH tunnel:
   ssh -N -f -L localhost:8888:turing.ce.pdn.ac.lk:8889 YOUR_USERNAME@tesla.ce.pdn.ac.lk
8. Open your browser and go to http://localhost:8888 . Paste the token when prompted.

---

Servers with public IP addresses (Tesla, Aiken)

These servers can be accessed directly. Tesla is used as an example here.

1. SSH into Tesla.
2. Create a conda environment and activate it.
   Read instructions
3. Install Jupyter: conda install jupyter
4. Start the notebook: jupyter notebook --no-browser --port=8080 --ip=0.0.0.0
5. You will see a URL like http://SOMETHING:8080/tree?token=abc123def456
6. Open your browser and go to http://tesla.ce.pdn.ac.lk:8080/tree?token=abc123def456 (replace the token with yours).

NOTE: If you are accessing Tesla from outside the university network, you will need to set up a SSH tunnel to Tesla to get access to a local port.