Upgrade your containerized deployment of Ansible Automation Platform

Perform an upgrade of containerized Ansible Automation Platform.

Before you begin

  • You have reviewed the release notes for the associated release. For more information, see Release notes.
  • You have a backup of your Ansible Automation Platform deployment. For more information, see Back up containerized Ansible Automation Platform.
  • You have provisioned the metrics_service database if using an external database. Metrics service requires access to two databases:
    • metrics_service database (read/write): Create a new database for metrics service data storage
    • awx/automation controller database (read-only): Metrics service requires read-only access to the existing controller database
Important:

Ansible Automation Platform 2.7 requires metrics service when automation controller is present. Review the requirements below to determine if you need to provision additional infrastructure:

  • Growth topology (all-in-one): No additional hardware required. Metrics service shares the existing host.
  • Multi-node deployments: Provision a new dedicated host for metrics service with the following specifications:
    • CPU: 2 vCPUs minimum, 4 vCPUs recommended
    • RAM: 4 GB minimum, 8 GB recommended
    • Storage: 20 GB minimum, 40 GB+ recommended (SSD preferred)

Procedure

  1. Log in to the Red Hat Enterprise Linux host as your dedicated non-root user.
  2. Follow the steps in Download Ansible Automation Platform to download the latest version of containerized Ansible Automation Platform.
  3. Copy the downloaded installation program to your Red Hat Enterprise Linux Host.
  4. Edit the inventory file to add the required [automationmetrics] group and update any other parameters to match your required configuration.

    You can keep the same parameters from your existing Ansible Automation Platform deployment, but you must add the [automationmetrics] inventory group.

    Important:

    The [automationmetrics] inventory group is required in Ansible Automation Platform 2.7 when [automationcontroller] is present. The installer will fail preflight checks if this group is missing.

    Example for growth topology (all-in-one):

    [automationmetrics]
aap.example.org

    Example for multi-node deployment:

    [automationmetrics]
metrics.example.org
  5. Add the following variables to the [all:vars] section: 
    # Metrics Service
automationmetrics_pg_host=<database_host>
automationmetrics_pg_database=metrics_service
automationmetrics_pg_username=metrics_service
automationmetrics_pg_password=<set your own>

# Read-only access to controller database
automationmetrics_controller_read_pg_host=<database_host>
automationmetrics_controller_read_pg_database=<controller_db_name>
automationmetrics_controller_read_pg_username=ms_awx_readonly
automationmetrics_controller_read_pg_password=<set your own>

    For complete inventory examples, see Container growth topology or Container enterprise topology.

  6. Run the install playbook: 
    $ ansible-playbook -i inventory ansible.containerized_installer.install
    • If your privilege escalation requires a password to be entered, append -K to the command. You will then be prompted for the BECOME password.
    • You can use increasing verbosity, up to 4 v's (-vvvv) to see the details of the installation process. However it is important to note that this can significantly increase installation time, so it is recommended that you use it only as needed or requested by Red Hat support.

Results

After successful upgrade, metrics service is running on the host defined in the [automationmetrics] inventory group.

Verify metrics service is operational:

  1. Check metrics service status by using the platform gateway.
  2. Review metrics service logs for any errors.