Troubleshoot Release Controller alerts

This section describes the investigation and troubleshooting steps for the Mirantis Container Cloud Release Controller service.

MCCReleaseControllerDeploymentStateCritical

Root cause

There are no Release Controller replicas scheduled in the cluster. By default, 3 replicas should be scheduled. The controller was either deleted or downscaled to 0.

Investigation

  1. Verify the status of the release-controller-release-controller deployment:

    kubectl get deployment -n kaas release-controller-release-controller

  2. Verify the .spec.replicas field value in the release-controller deployment spec:

    kubectl get deployment -n kaas \
release-controller-release-controller -o=json | jq -r \
'.spec.replicas'

Mitigation

If the Release Controller deployment has been downscaled to 0, set the replicas back to 3 in the release-controller Helm release in the .spec.replicas section of the Deployment object on the management cluster:

kubectl edit deployment -n kaas release-controller-release-controller