Create an Anaconda distribution instance with a conda environment

In this lesson, you create and deploy a default Anaconda distribution and add a conda environment.

About this task

IBM® Spectrum Conductor offers better integration of Anaconda with Spark workload and Jupyter notebooks that includes enterprise security. Start by adding an Anaconda distribution in the cluster management console. This can be either a new or existing one. Next you deploy the Anaconda distribution, which creates an Anaconda distribution instance. Then you add a conda environment that contains all of the packages the Anaconda distribution instance requires. You can either specify a conda environment name to create an environment with either all standard Python libraries or no packages; or you can drag or select an Anaconda environment file to create an environment based on a yaml file.

This lesson uses the following concepts:
Concept Description
Anaconda distribution An Anaconda distribution is a Python/R data science distribution that installs 1,400+ data science packages for Python/R, with the ability to manage packages, dependencies, and environments.
Anaconda distribution instance A deployment of an Anaconda distribution in IBM Spectrum Conductor.
conda environment A directory that contains a specific collection of conda packages. Within a conda environment, you can have various collections of conda packages. A conda package is a compressed tarball file that contains system-level libraries, Python or other modules, executable programs, and other components. Conda keeps track of the dependencies between packages and platforms.

Procedure

  1. From the cluster management console, click Resources > Frameworks > Anaconda Management, and select the default distribution that matches your operating system.
  2. Click Deploy.
  3. On the Deployment Settings tab enter the following information:
    • Instance name: sampleanaconda
    • Deployment directory: /home/LOBExecUser/sampleanaconda
    • Consumer: / (Root Consumer)
    • Resource group: All Hosts
    • Execution user: LOBExecUser
  4. Click Deploy.
    The Anaconda distribution begins deploying.

    Click Continue to Anaconda Distribution Instance to view the status of the distribution. Look at the deployment state section the page to see the status. For a deployed distribution, it should be in the Ready state. If you see Deploy error, click the distribution to view the deployment error.

  5. Under Conda Environments, click Add.
  6. Clear Create environment from a yaml file.
  7. Leave Clone base environment selected to install all standard Python libraries. If you clear the check box, you have to manually add them.
  8. For Environment name, type in sampleenv.
  9. Click Add > Close.

    Once a conda environment is added, the installed conda packages are listed when you click the environment name.

Results

You now have an deployed Anaconda distribution instance with a conda environment in your cluster.

Summary

In this lesson, you learned how to create and deploy a default Anaconda distribution and add a conda environment. This lesson concludes the resource planning module.

In the next module, you will learn about Spark workload.