Uploaded image for project: 'OpenShift Request For Enhancement'
  1. OpenShift Request For Enhancement
  2. RFE-2845

Make openshift-install check if all worker nodes are healthy

XMLWordPrintable

    • False
    • None
    • False
    • Not Selected
    • 0
    • 0% 0%

      1. Proposed title of this feature request

      Make openshift-install check if all worker nodes are healthy using the compute.replicas field of install-config.yaml as reference, and make it return 0 only if this check is successful

      2. What is the nature and description of the request?

      At the moment, in order to assess the health of a new cluster, openshift-install checks the status of the api object ClusterVersion,which represents a "summary" of all the ClusterOperator objects statuses; however ClusterOperators  only become degraded when they can't perform their own functions; that is, a unhealthy worker node doesn't necessarily constitute a failure.

      This cause openshift-install to return 0 (that is, report a successful installation) even if some worker node is still unhealthy (for example, see situation described on [this KCS|https://access.redhat.com/solutions/6957136])

      This proposal is to add an extra check in openshift-install in order to make it assess if the number of Ready  worker nodes is equal to the number in the compute.replicas field of install-config.yaml,and make {{openshift-install return 0 only if this check is successful.

      3. Why does the customer need this? (List the business requirements here)

      The customer is using openshift-install  in a automated pipeline which deploys a massive number of clusters, and is relying on the openshift-install return code to verify that all the cluster components are ok.

      4. List any affected packages or components.

      openshift-install 

       

            rhn-support-dhardie Duncan Hardie
            rhn-support-acancell Alfonso Cancellara
            Ivan Bodunov
            Votes:
            0 Vote for this issue
            Watchers:
            6 Start watching this issue

              Created:
              Updated:
              Resolved: