Run Jupyter Notebook on a GPU Server

Prerequisites

• A running LightYear GPU server
• Python 3.10+ installed
• Port 8888 open in your firewall group

Step 1 — Install Jupyter

$pip3 install jupyterlab

Step 2 — Generate a Config File

$jupyter lab --generate-config

Step 3 — Set a Password

$jupyter lab password
$# Enter and confirm a password

Step 4 — Configure Remote Access

Edit ~/.jupyter/jupyter_lab_config.py:

c.ServerApp.ip = '0.0.0.0'
c.ServerApp.port = 8888
c.ServerApp.open_browser = False
c.ServerApp.allow_root = True

Step 5 — Open the Firewall

In the LightYear control panel, add a firewall rule:

Step 6 — Start Jupyter Lab

$jupyter lab --no-browser &

Access at: http://<SERVER_IP>:8888

Running as a systemd Service

# /etc/systemd/system/jupyter.service
[Unit]
Description=Jupyter Lab
After=network.target

[Service]
Type=simple
User=root
ExecStart=/usr/local/bin/jupyter lab --no-browser
Restart=on-failure

[Install]
WantedBy=multi-user.target

$systemctl daemon-reload
$systemctl enable jupyter
$systemctl start jupyter

Using SSH Tunnelling (More Secure)

Instead of opening port 8888 publicly, tunnel via SSH:

$ssh -L 8888:localhost:8888 root@<SERVER_IP>

Then access http://localhost:8888 on your local machine.