Connecting to Linux via Jupyter Notebook on Windows Server - jamiefogel/Networks GitHub Wiki

Step 1: Set Up the SSH Tunnel with PuTTY

1. Open PuTTY and configure the connection to your Linux server.

   • Under Session, enter the hostname or IP address of your Linux server and the port (usually 22).

2. Configure Port Forwarding in PuTTY:

   • In PuTTY, go to Connection > SSH > Tunnels.
   • Set up local port forwarding by entering a local port, such as 8888, in the Source port field.
   • In the Destination field, enter the address and port of the Jupyter server on your Linux machine (e.g., localhost:8888 if you start Jupyter on that port).
   • Click Add to confirm the tunnel.

3. Start the SSH Connection:

   • Go back to Session and click Open to initiate the SSH connection. This will forward traffic from localhost:8888 on your Windows machine to the server’s localhost:8888.

Note: I have saved this configuration in PuTTY as 10.1.1.226 - Jupyter Notebook so at this point we can just click on that. This will open a terminal and we enter out username and password.

Step 2: Start a Jupyter Notebook on the Linux Server (Optional for Spyder Kernel)

First activate our environment by entering in the PuTTY terminal

conda activate labor_gt 

On the Linux server, you can start a Jupyter Notebook server to use as a remote kernel, which can be accessed through the SSH tunnel:

jupyter notebook

This will print a bunch of text in the terminal. Look for the text To access the server, open this file in a browser:. This will be followed by a few options. The one that has seemed to work is the one that begins with http://graphtool-prod:. Copy this text and paste it as a url into a web browser and this should open our home folder in a Jupyter