3.6

SSH Server Feature

Enterprise Applicable to Enterprise Edition
Community Applicable to Community Edition

PrimeHub v3.0+ provides a new feature of SSH Server that users are allowed to access into their own launched JupyterHub via SSH. A practical use-case is that users are able to use IDEs such as VSCode which has remote-ssh extension for the development remotely. In other words, it provides other possible development options for users who like to use rich-features IDEs instead.

Enable SSH Server

To activate SSH Server, administrators have to configure SSH Server of PrimeHub. Please contact administrators and see Configure SSH Server.

Before using this feature, you are required to have a public/private key pair, please go to SSH key generation to generate your own pair at local.

  1. Enter JupyterHub from side-menu, before staring a notebook, expand Show advanced settings and check Enable SSH Server.

  2. Hovering the cursor over the code block, copy the configuration by clicking the copy icon.

  3. Press Start Notebook to launch the JupyterHub.

  4. Go back to your local, edit the ~/.ssh/config with adding the copied configuration at your local. Make sure jupyter-<username> and ~/.ssh/id_rsa(private key) are correct according to your circumstance.

    If you want to assign other private key, please change the path ~/.ssh/id_rsa in the configuration to the desired one. If your key requires passphrase, you may want to add into ssh-agent.

Deploy Public key into JupyterHub

After JupyterHub is launched,

In JupyterHub menu on User Portal, it shows the same instruction for the reference.

  1. Drag-n-drop a public key (e.g. id_rsa.pub) from your local to the File Browser of the opened JupyterHub tab on browser.

  2. Open Terminal on JupyterHub, run the commands

    mkdir -p ~/.ssh
mv ~/id_rsa.pub ~/.ssh/authorized_keys

  3. Go back to your local, run the command for verification.

    ssh jupyter

    After SSH into the JupyterHub successfully, you should see the prompt:

    jovyan@jupyter-<your_username>:~$

Once the initial setup is completed, only step 1 is required for accessing JupyterHub via SSH.

Practical Usecase - VSCode

If you are familiar with VSCode, check our guide, VSCode + Remote-SSH.

Troubleshooting

If permission denied or connection refused occurs, please check our advices for troubleshooting.

← Advanced SettingsTips →