Chapter 12. Configuring Ansible Tower tasks to run on managed clusters

Select the appropriate path for your situation:

Select the credential that you want to use for the new template. To link an Ansible credential to an Ansible template, complete the following steps:

Select the prehook and posthook Ansible jobs to add to the installation or upgrade of the cluster.

Note: The Ansible job template name must match the name of the Ansible job on the Ansible Tower.

Create the definition file for the Ansible job in one of the channels that are supported by the application function. Only Git channels are supported.

Use AnsibleJob as the value of kind in the definition.

Your definition file contents might resemble the following example:

apiVersion: apiVersion: tower.ansible.com/v1alpha1
kind: AnsibleJob
metadata:
  name: hive-cluster-gitrepo
spec:
  tower_auth_secret: my-toweraccess
  job_template_name: my-tower-template-name
  extra_vars:
    variable1: value1
    variable2: value2

By storing the file in the prehook or posthook directory, it creates a list of cluster names that match the placement rule. The list of cluster names can be passed as a value of extra_vars to the AnsibleJob kind resource. When this value is passed to the AnsibleJob resource, the Ansible job can determine the new cluster name and use it in the automation.

Using the Red Hat Advanced Cluster Management console, create an application with a Git subscription that references the channel where you stored the definition file that you just created. See Managing application resources for more information about creating an application and subscription.

When you create the subscription, specify a label that you can add to the cluster that you create or import later to connect this subscription with the cluster. This can be an existing label, like vendor=OpenShift, or a unique label that you create and define.

Note: If you select a label that is already in use, the Ansible job automatically runs. It is best practice to include a resource in your application that is not part of the prehooks or posthooks.

The default placement rule runs the job when it detects the cluster with the label that matches the label of the AnsibleJob. If you want the automation to run on all of your running clusters that are managed by the hub cluster, add the following content to the placement rule:

clusterConditions:
  - type: ManagedClusterConditionAvailable
    status: "True"

You can either paste this into the YAML content of the placement rule or you can select the option to Deploy to all online clusters and local cluster on the Application create page of the Red Hat Advanced Cluster Management console.

Create or import your cluster by following the instructions in Creating a cluster or Importing a target managed cluster to the hub cluster, respectively.

When you create or import the cluster, use the same label that you used when you created the subscription, and the AnsibleJob is automatically configured to run on the cluster.

Create the definition file for the AnsibleJob in the prehook or posthook directory of your Git Channel.

Use AnsibleJob as the value of kind in the definition.

Your definition file contents might resemble the following example:

apiVersion: tower.ansible.com/v1alpha1
kind: AnsibleJob
metadata:
  name: hive-cluster-gitrepo
spec:
  tower_auth_secret: my-toweraccess
  job_template_name: my-tower-template-name
  extra_vars:
    variable1: value1
    variable2: value2

Replace my-toweraccess with the authentication secret to access your Ansible Tower.

Replace my-tower-template-name with the template name from your Ansible Tower.

View the status of the last run of the Ansible job on the cluster information. The entry shows one of the following statuses: