Install and Configure COO on CRC

Install the Red Hat OpenShift Cluster Observability Operator (COO) on an OpenShift Local (CRC) cluster and enable Perses dashboards and incident detection via the monitoring UIPlugin.

Prerequisites

The CRC cluster is running and you are logged in as kubeadmin.
Cluster monitoring must be enabled. If it is not, use the crc-cluster-monitoring skill first — COO and incident detection depend on the monitoring stack being active.

Steps

Verify cluster monitoring is running Confirm that the monitoring stack is available before proceeding:
```
oc get co monitoring
```
The output must show Available=True and Degraded=False. If the monitoring ClusterOperator does not exist or is degraded, stop and enable cluster monitoring first.

Create the COO namespace with the cluster-monitoring label The openshift.io/cluster-monitoring: "true" label is required for incident detection to work:

cat <<'EOF' | oc apply -f -
apiVersion: v1
kind: Namespace
metadata:
  name: openshift-cluster-observability-operator
  labels:
    openshift.io/cluster-monitoring: "true"
EOF

Create the OperatorGroup and Subscription Use the supporting script to install COO from the redhat-operators catalog:
```
bash <skill-dir>/install-coo.sh
```
Wait for the COO operator to be ready Wait for the CSV to reach Succeeded and the operator pods to start:
```
oc wait --for=jsonpath='{.status.phase}'=Succeeded \
  csv -l operators.coreos.com/cluster-observability-operator.openshift-cluster-observability-operator \
  -n openshift-cluster-observability-operator --timeout=300s
```
Then confirm the pods are running:
```
oc get pods -n openshift-cluster-observability-operator
```
You should see at least the following pods running:
- observability-operator-*
- obo-prometheus-operator-*
- obo-prometheus-operator-admission-webhook-* (2 replicas)
- perses-operator-*

Create the UIPlugin CR with Perses and incident detection Apply the UIPlugin custom resource to enable both features:

cat <<'EOF' | oc apply -f -
apiVersion: observability.openshift.io/v1alpha1
kind: UIPlugin
metadata:
  name: monitoring
spec:
  type: Monitoring
  monitoring:
    perses:
      enabled: true
    clusterHealthAnalyzer:
      enabled: true
EOF

Wait for the UIPlugin to reconcile

oc wait --for=jsonpath='{.status.conditions[0].reason}'=UIPluginReconciled \
  uiplugin monitoring --timeout=120s

Verification

oc get csv -n openshift-cluster-observability-operator shows the COO CSV in Succeeded phase
oc get pods -n openshift-cluster-observability-operator shows all pods Running (including perses-operator, perses-0, and cluster-health-analyzer)
oc get crd | grep perses returns all 4 Perses CRDs (perses, persesdashboards, persesdatasources, persesglobaldatasources)
oc get uiplugin monitoring -o jsonpath='{.status.conditions[?(@.type=="Available")].status}' returns True
oc get uiplugin monitoring -o jsonpath='{.status.conditions[?(@.type=="Degraded")].status}' returns False
oc get consoleplugin lists monitoring-console-plugin
The OpenShift web console shows Observe → Incidents (may take ~5 minutes for timeline data to appear)
The OpenShift web console shows Observe → Dashboards for Perses

Notes

After the UIPlugin is created, the web console may reload automatically. If the new menu items do not appear, refresh the browser.
Incident detection takes approximately 5 minutes to begin correlating alerts after enablement. Only alerts firing after enablement are analyzed.
The Perses Operator is installed automatically with COO. You do not need to install it separately. The UIPlugin CR creates the Perses server instance and registers the console plugin.
To use Perses dashboards with cluster metrics, you will need to create a PersesGlobalDatasource pointing to the platform Thanos Querier.

Changelog

See CHANGELOG.md for version history.