remote_jupyter.md

Remote Jupyter

Easily run Jupyter Notebook on an EC2 (say), and connect to a localhost port on your local machine.

ssh into the remote

We will use localhost:8888 on the local machine, and 8889 on the remote. On your local machine (using your own public DNS):

    ssh -i ~/.ssh/mykey.pem -L localhost:8888:localhost:8889 ubuntu@12-34-56-78.compute-1.amazonaws.com

This is going to open port 8888 locally to SSH traffic from port 8889 on the remote.

Note that this won't work if port 8888 on your local is already bound to some other service. If it fails, change it to some other number, eg 8899.

Note also that adding the flags -N -f will provide a non-interactive session that goes into the background. Handy for when you're already connected in a terminal.

Get ready on the remote machine

• Set up the remote machine, as usual. Install Anaconda (go here and get the link to the latest version):

    wget https://repo.continuum.io/archive/Anaconda3-4.4.0-Linux-x86_64.sh
    bash ./Anaconda3-4.4.0-Linux-x86_64.sh

• Create your environment:

    conda create -n myenv anaconda
    source activate myenv

• Install anything else you need.

Get things ready on the local machine

• Start Jupyter, e.g. on port 8889 (default is 8888, I just changed it to illustrate):

    jupyter notebook --no-browser --port=8889

Copy the link and paste it into your browser. You're done.

Some resources in case you run into trouble

• https://help.ubuntu.com/community/SSH/OpenSSH/PortForwarding
• https://coderwall.com/p/ohk6cg/remote-access-to-ipython-notebooks-via-ssh
• https://manas.tungare.name/blog/ssh-port-forwarding-on-mac-os-x/