Enable automation dashboard during operator installation

Enable automation dashboard using declarative Custom Resource configuration to leverage operator-managed dashboard deployment with automatic feature flag propagation to metrics service.

Before you begin

  • AnsibleAutomationPlatform CR has been deployed and reconciled successfully
  • Metrics service enabled in the AnsibleAutomationPlatform CR (spec.metrics section present with disabled: false)
  • You are running Ansible Automaton Platform operator (aap-operator) and metrics service operator (automationmetricsservice-operator) running
  • Understanding of Technology Preview features and limitations
Important

Technology Preview: Automation dashboard is a Technology Preview feature in Red Hat Ansible Automation Platform 2.7 and is disabled by default. You must explicitly enable it by setting spec.feature_flags.FEATURE_DASHBOARD_COLLECTION_ENABLED: true in your AnsibleAutomationPlatform Custom Resource.

About this task

This procedure enables automation dashboard for Red Hat Ansible Automation Platform 2.7 deployments using the operator on Kubernetes or OpenShift. By configuring the AnsibleAutomationPlatform Custom Resource with the dashboard feature flag, the operator automatically propagates the configuration to metrics service within the reconciliation cycle (typically under 2 minutes). This declarative approach eliminates manual pod configuration and enables version-controlled, GitOps-compatible dashboard state management with 100% operator-managed deployment.

Procedure

  1. Edit the AnsibleAutomationPlatform Custom Resource

    Edit your existing AnsibleAutomationPlatform CR to enable the dashboard collection feature flag. The feature flag is set at the top level of spec, not inside spec.metrics.

    kubectl edit AnsibleAutomationPlatform <aap-cr-name> -n <namespace>

    Or, if using OpenShift:

    oc edit AnsibleAutomationPlatform <aap-cr-name> -n <namespace>

    Add or update the feature_flags section:

    spec:
  feature_flags:
    FEATURE_DASHBOARD_COLLECTION_ENABLED: true   # ADD THIS
  metrics:
    disabled: false
    name: <aap-cr-name>-metrics
  # ... rest of your existing spec unchanged
    Expand
    Field Default Purpose
    spec.feature_flags.FEATURE_DASHBOARD_COLLECTION_ENABLED false (absent) Enables automation dashboard data collection in metrics service (Technology Preview)
    spec.metrics.disabled false Enables the metrics service component. Must befalse for dashboard to function.
    Note

    spec.feature_flags is a top-level spec field on the AnsibleAutomationPlatform CR. Do not nest it under spec.metrics.

    Save and exit. The Ansible Automation Platform operator automatically begins reconciliation.

  2. Verify Ansible Automation Platform operator reconciliation

    The Ansible Automation Platform operator reconciles the AnsibleAutomationPlatform CR and propagates FEATURE_DASHBOARD_COLLECTION_ENABLED to the MetricsService CR. Check the Ansible Automation Platform operator logs:

    kubectl logs -n <namespace> \
deployment/aap-gateway-operator-controller-manager | tail -20

    Look for reconciliation activity on your Ansible Automation Platform CR name.

  3. Verify the feature flag reached the metrics service

    Expected output:

    METRICS_SERVICE_FEATURE_DASHBOARD_COLLECTION_ENABLED: '@bool True'

    If this key is absent, the operator has not yet reconciled. Wait 1–2 minutes and re-check.

    Note

    The @bool True prefix is Dynaconf syntax for boolean values. This is expected and correct.

    You can also verify by using the pod environment directly:

    METRICS_POD=$(kubectl get pods -n <namespace> \
  -l app.kubernetes.io/component=metrics-service \
  --field-selector=status.phase=Running \
  -o jsonpath='{.items[0].metadata.name}')

kubectl exec -n <namespace> $METRICS_POD -- \
  env | grep DASHBOARD

    Expected output:

    METRICS_SERVICE_FEATURE_DASHBOARD_COLLECTION_ENABLED=@bool True
  4. Verify metrics service pods restarted

    After the ConfigMap is updated, the metrics service pods restart automatically to pick up the new environment variable. Confirm three pods are running:

    kubectl get pods -n <namespace> | grep metrics

    Expected output (on three running pods):

    <aap-cr-name>-metrics-web-<hash>        1/1   Running   0   <age>
<aap-cr-name>-metrics-tasks-<hash>      1/1   Running   0   <age>
<aap-cr-name>-metrics-scheduler-<hash>  1/1   Running   0   <age>
    Note

    There is no separate settings.yaml file in the operator deployment. All metrics service configuration is managed by using the <name>-metrics-env-properties ConfigMap.
  5. Verify dashboard tables created 
    # Get database pod
DB_POD=$(kubectl get pods -n <namespace> | grep postgres | awk '{print $1}')

# Check dashboard tables exist
kubectl exec -n <namespace> $DB_POD -- \
  psql -U metrics_service -d metrics_service \
  -c "SELECT table_name FROM information_schema.tables WHERE table_name LIKE 'dashboard%';"

    Expected output: Six dashboard tables:

    • dashboard_job_data
    • dashboard_job_data_host_summary
    • dashboard_job_data_label
    • dashboard_template_metadata
    • dashboard_subscription_cost
    • dashboard_filter_set

Results

Dashboard is successfully enabled when:

  • spec.feature_flags.FEATURE_DASHBOARD_COLLECTION_ENABLED: true is present in the AnsibleAutomationPlatform CR
  • <name>-metrics-env-properties ConfigMap contains METRICS_SERVICE_FEATURE_DASHBOARD_COLLECTION_ENABLED: '@bool True'
  • Three metrics service pods running (web, tasks, scheduler)
  • Six dashboard tables exist in metrics_service database

Troubleshoot

Issue: Feature flag not appearing in ConfigMap

Symptom: After editing the Ansible Automation Platform CR, METRICS_SERVICE_FEATURE_DASHBOARD_COLLECTION_ENABLED does not appear in the ConfigMap.

Solution:

  1. Confirm the change was applied to the Ansible Automation Platform CR:

    kubectl get AnsibleAutomationPlatform <aap-cr-name> \
  -n <namespace> -o yaml | grep -A3 feature_flags

    Verify FEATURE_DASHBOARD_COLLECTION_ENABLED: true is present under spec.feature_flags.

  2. Check Ansible Automation Platform operator logs for reconciliation errors:

    kubectl logs -n <namespace> \
  deployment/aap-gateway-operator-controller-manager | grep -i error

  3. Check automationmetricsservice operator logs:

    kubectl logs -n <namespace> \
  deployment/automationmetricsservice-operator-controller-manager | tail -30

Dashboard UI returns 403 Forbidden errors

Symptom: Dashboard navigation appears in AAP UI but clicking it shows authentication errors

Cause: Gateway has not registered metrics service for routing and authentication (known issue in early builds)