Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-29110

HCP: imageRegistryOverrides information are extracted only once on HyperShift operator initialization and never refreshed

XMLWordPrintable

    • No
    • Hypershift Sprint 249, Hypershift Sprint 252, Hypershift Sprint 253
    • 3
    • Rejected
    • False
    • Hide

      None

      Show
      None
    • Hide
      * With this update, in disconnected deployments, the HyperShift Operator receives the new ICSP/IDMS from the management cluster and adds them to the HyperShift Operator and the Control Plane Operator in every reconciliation loop. The changes to the ICSP/IDMS cause the control-plane-operator pod to be restarted.(link:https://issues.redhat.com/browse/OCPBUGS-29110[*OCPBUGS-29110*])
      Show
      * With this update, in disconnected deployments, the HyperShift Operator receives the new ICSP/IDMS from the management cluster and adds them to the HyperShift Operator and the Control Plane Operator in every reconciliation loop. The changes to the ICSP/IDMS cause the control-plane-operator pod to be restarted.(link: https://issues.redhat.com/browse/OCPBUGS-29110 [* OCPBUGS-29110 *])
    • Bug Fix
    • Done

      Description of problem:

      HO uses the ICSP/IDMS from mgmt cluster to extract the OCP release metadata to be used in the HostedCluster.

But they are extracted only once in main.go:
https://github.com/jparrill/hypershift/blob/9bf1403ae09c0f262ebfe006267e3b442cc70149/hypershift-operator/main.go#L287-L293
before starting the HC and NP controllers but they are never refreshed anymore when ICSP/IDMS changes on the management cluster neither when a new HostedCluster is created.

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

          4.14 4.15 4.16

      How reproducible:

      100%

      Steps to Reproduce:

          1. ensure that HO is already running
    2. create an ICSP or a IMDS on the management cluster
    3. try to create an hosted-cluster

      Actual results:

      the imageRegistryOverrides setting for the new hosted-cluster ignores the ICSP/IMDS created when the HO was already running.
Killing HO operator pod and wait for it to restart will bring to a different result.

      Expected results:

      HO is consistently consuming  ICSP/IMDS info at runtime without the need to be restarted

      Additional info:

          It affects disconnected deployments

              jparrill@redhat.com Juan Manuel Parrilla Madrid
              stirabos Simone Tiraboschi
              He Liu He Liu
              Votes:
              0 Vote for this issue
              Watchers:
              9 Start watching this issue

                Created:
                Updated:
                Resolved: