Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-51152

IPI installation on VMWare not possible like documented with blank failureDomains

    • Quality / Stability / Reliability
    • False
    • Hide

      None

      Show
      None
    • None
    • Moderate
    • Yes
    • None
    • None
    • None
    • OSDOCS Sprint 284
    • 1
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      Description of problem:

      Installing a cluster without a failure domain is not possible. The [documentation|https://docs.openshift.com/container-platform/4.17/installing/installing_vsphere/ipi/installing-vsphere-installer-provisioned-customizations.html#installation-vsphere-regions-zones_installing-vsphere-installer-provisioned-customizations] is mentioning the following:
      
      The default install-config.yaml file includes vcenters and failureDomains fields, where you can specify multiple vSphere data centers and clusters for your OpenShift Container Platform cluster. You can leave these fields blank if you want to install an OpenShift Container Platform cluster in a vSphere environment that consists of single data center.
          

      Version-Release number of selected component (if applicable):

      openshift-installer 4.13+
          

      How reproducible:

      Always
          

      Steps to Reproduce:

          1. create install-config.yaml with blank failureDomains
          2. install cluster
          

      Actual results:

      The installer is failing with an error:
      failed to fetch Metadata: failed to load asset \"Install Config\": failed to create install config: failed to unmarshal install-config.yaml: error unmarshaling JSON: while decoding JSON: json: cannot unmarshal object into Go struct field Platform.platform.vsphere.failureDomains of type []vsphere.FailureDomain"
          

      Expected results:

      installation should be possible for single data center environments without specifying a failureDomains
          

      Additional info:

      A customer which is managing their existing environment with a couple of clusters, can't just specify a pseudo failureDomain as a workaround as the management of the ControlPlaneMachineSet without a failureDomain (which manged by GitOps) is breaking by the admission webhook "controlplanemachineset.machine.openshift.io".
      There should be a way to add a failuredomain to existing cluster or install a cluster without a failuredomain to make sure the environment stay consistent.
      
          

              rhn-support-sgehlot Sandeep Gehlot
              rhn-support-anowak Andreas Nowak
              None
              None
              Shang Gao Shang Gao
              None
              Votes:
              0 Vote for this issue
              Watchers:
              6 Start watching this issue

                Created:
                Updated:
                Resolved: