Empty IBMCOS storage config causes operator to crashloop

[hasueki]

Empty IBMCOS storage config causes operator to crashloop

[bparees]

can you file a bug against the Image Registry component here?

https://issues.redhat.com/secure/CreateIssueDetails!init.jspa?pid=12332330&issuetype=1&summary=Bug+summary+here&description=Describe+what+you+were+doing+and+the+behavior+you+saw

[hasueki]

Corresponding bug report: https://issues.redhat.com/browse/OCPBUGS-6797

[hasueki]

The imageregistry config is generated in this bad format when the Infrastructure object for the cluster is not coming from IPI clusters (i.e. IBM ROKS, HyperShift on IBM Cloud):

apiVersion: config.openshift.io/v1
kind: Infrastructure
metadata:
  labels:
    hypershift.openshift.io/managed: "true"
  name: cluster
spec:
  cloudConfig:
    name: ""
  platformSpec:
    type: IBMCloud
status:
  apiServerInternalURI: https://h6b4f4a1443ee912a989e-d603ff82e51c94176a53d44566df9d79-c000.us-south.satellite.test.appdomain.cloud:32696
  apiServerURL: https://h6b4f4a1443ee912a989e-d603ff82e51c94176a53d44566df9d79-c000.us-south.satellite.test.appdomain.cloud:32696
  controlPlaneTopology: External
  etcdDiscoveryDomain: cfbl2p710movqf6tdos0.hide-prestage-co-9e37478581b5d9de33607f5926d1d18f-0000.us-south.prestg.stg.containers.appdomain.cloud
  infrastructureName: cfbl2p710movqf6tdos0
  infrastructureTopology: HighlyAvailable
  platform: IBMCloud
  platformStatus:
    type: IBMCloud

We could potentially leverage the controlPlaneTopology: External to toggle different configs for such clusters.