Skip to main content

Home

Installing Unravel on cloud platforms using Ansible

You can install Unravel on Amazon EMR, Databricks (Azure, AWS), and Google cloud platform (Dataproc, BigQuery) using Ansible.

  1. Navigate to the ansible-installation directory and create a copy of the vars_template.yml file.

    cp vars_template.yml vars.yml
  2. Edit vars.yml by using any editor. For example, the vi editor.

    vi vars.yml
  3. Update the value for the cloud_platform parameter.

    # Update any required fields
    # Must set cloud_platform=emr/dataproc/databricks

    The vars.yml file contains deployment parameters. You can customize these parameters for installing and deploying Unravel on cloud platforms.

  4. From the Unravel server, run the Ansible playbook.

    ansible-playbook -i inventories/ install_unravel.yml -e @vars.yml -vvv

    Applies all values specified in the vars.yml to the Ansible playbook. See Verify Unravel installation using Ansible.

  5. (Optional) Update the cloud_node inventory if you do not run the Ansible playbook from the Unravel server.

    Command

    vi inventories/cloud_node/hosts
    

    Output

    # Example hosts file
    # [cloud_node]
    # abc.unraveldata.com ansible_user=unravel ansible_ssh_private_key_file=/root/.ssh/id_rsa
  6. Define the hive metastore properties in a .txt file for Databricks, EMR, and Dataproc. See Configuring Hive Metastore (Cloud) and Configuring external hive metastore (EMR).

  7. Specify the BigQuery configuration in Unravel. See Add BigQuery details in Unravel.

  8. Configure the EMR cluster in Unravel. See Connect a new or existing EMR cluster to Unravel.

  9. Configure the Dataproc cluster in Unravel. See Connecting Unravel Server to a new Dataproc cluster.

  10. Specify the path of a .txt file in the prop_file_paths variable. Ensure that the .txt file is readable by the Unravel user.

    ansible-playbook -i inventories/ install_unravel.yml -e @vars.yml -vvv --tag hive_metastore