In cluster-ingress-operator's ensureNodePortService, when there is a conflicting Ingress Controller loadbalancer, it states:

a conflicting load balancer service exists that is not owned by the ingress controller: openshift-ingress/router-loadbalancer

Technically that is the service name, not the ingress controller name. The IC name is openshift-ingress/loadbalancer in this example.

So the error message wording is incorrect.

4.13
4.12
4.11

Easy

# Create a service that will conflict with a new ingress controller
oc create svc nodeport router-nodeport-test --tcp=80 -n openshift-ingress
DOMAIN=$(oc get ingresses.config/cluster -o jsonpath={.spec.domain})
oc apply -f - <<EOF
apiVersion: operator.openshift.io/v1
kind: IngressController
metadata:
  name: test
  namespace: openshift-ingress-operator
spec:
  domain: reproducer.$DOMAIN
  endpointPublishingStrategy:
    type: NodePortService
  replicas: 1
  nodePlacement:
    nodeSelector:
      matchLabels:
        node-role.kubernetes.io/worker: ""
EOF

# Look for log message that is incorrect
oc logs -n openshift-ingress-operator $(oc get -n openshift-ingress-operator pods --no-headers | head -1 | awk '{print $1}') -c ingress-operator | grep conflicting 

# The results provide service name, not ingress controller name
# "error": "a conflicting nodeport service exists that is not owned by the ingress controller: openshift-ingress/router-test"  

"error": "a conflicting nodeport service exists that is not owned by the ingress controller: openshift-ingress/router-test"

"error": "a conflicting nodeport service exists that is not owned by the ingress controller: openshift-ingress/router-nodeport-test"