Installing Jupyter Notebook Server on Linux

You can install IBM® Cognos® Analytics for Jupyter Notebook Server either on the same computer or on a different computer from where Cognos Analytics is installed.

The Jupyter Notebook Server supports Linux and Windows 10 platforms and requires Docker to be installed.

Note: Docker CE (Community Edition), Docker Engine, and Docker Desktop (CE) are supported at this time.

When you download and run the installer script, you load and start Docker containers. These containers allow Cognos Analytics users to create and edit Jupyter Notebook. By default, Cognos Analytics for Jupyter Server is configured with many of the most common data science/analytic Python packages.

Tip: You can later upgrade the Python packages in your existing installation.

Before you begin

Before you install Jupyter Notebook Server, follow these steps:

  1. Install Docker or Podman (on RHEL 8 or RHEL 9) as the container engine for Jupyter Notebook.
    To install Docker:
    Follow the procedures for one of these Linux distributions:

    You must be added to the Docker group for any Docker command to run without root privileges.

    To install Podman on RHEL8 or RHEL9:
    Follow the installation and configuration steps at Podman.
    Note: Throughout the Podman documentation, you will see Podman commands which correspond to particular topics where Docker commands were available.
    One of Podman’s greatest advantages is its complete CLI compatibility with Docker. When building Podman, Docker users can adapt without any significant changes. For example, you can use the alias command to create a docker alias for Podman: 
    yum install podman-docker
    Important: To enable advanced networking features in Jupyter Server, you must also install the package podman-plugins. Type the following:
    yum install -y podman-plugins
    For more information on Podman, see this Podman developer blog.
  2. Set the fully qualified domain in Cognos Configuration.
    1. In IBM Cognos Configuration, in the Explorer window, click Environment.
    2. Under the Dispatcher URI for external applications set the fully qualified domain name (FQDN) for the IBM Cognos Analytics server.
    3. Under the Gateway URI set the fully qualified domain name (FQDN) for the IBM Cognos Analytics server.
    4. Click File > Save.
    5. Restart the Cognos Analytics service.

Procedure

  1. Download the IBM Cognos Analytics for Jupyter Notebook Server installer and Server repository from Passport Advantage.
    Tip: The Jupyter Server installer and Server repository are found only in the Linux eAssemblies.
  2. Double click the installer file.
  3. Follow the directions in the installation wizard to copy and install the files to your computer.
    Tip: You can install on top of an older version of Jupyter Server.
    The folder jupyter_installation_location/dist folder contains two subfolders:
    • dist/images
    • dist/scripts
      Tip: The folder dist/scripts/unix contains all the scripts that you need to run.

      Script

      		 Purpose
      build.sh Run this to rebuild the images.
      config.conf Edit this configuration file to change Jupyter parameters.
      install.sh Run this to load and start the Docker containers.
      prune.sh Run this to remove old Docker images.
      start.sh Run this to start the Jupyter server.
      stop.sh Run this to stop the Jupyter server.
      uninstall.sh Run this to uninstall Jupyter server.
  4. Ensure that you have execute permissions for each script:

    Type chmod -R u+x dist/scripts/unix

  5. Navigate to the dist/scripts/unix directory.
  6. Type ./install.sh

    The install script runs.

Results

All of the Jupyter Server Docker images are loaded from the jupyter_installation_location/dist/images directory and the Docker containers are started.

What to do next

After installing IBM Cognos Analytics for Jupyter Notebook Server, the following tasks can be performed:

Uninstalling Jupyter Notebook Server

To uninstall Jupyter Notebook Server, navigate to the dist/scripts/unix directory and then type ./uninstall.sh