Description of problem:

The `grafana-dashboard-cluster-logging` cm is not recreated when the Cluster Logging Operator is installed in Single Namespace, this means, only in `openshift-logging` leading to don't be visible the Dashboard `Logging / Collection"

Then, an user installing Logging before 5.8 and upgrading until 5.9.0 hit the bug https://issues.redhat.com/browse/LOG-5308. With this bug, the "Logging / Collection" was removed and even with the patch, if the Operator Group continues as single namespace:

apiVersion: operators.coreos.com/v1
kind: OperatorGroup
metadata:
  name: cluster-logging
  namespace: openshift-logging 
spec:
  targetNamespaces:
    - openshift-logging

Then, the Dashboard is not watched by the Cluster Logging Operator and never recreated

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

4.14
CLO 5.8.13 and 5.9.7 where z is the latest version available now

How reproducible:

Always

Steps to Reproduce:

1. Install Cluster Logging 5.8 or 5.9 in latest version in single namespace

   apiVersion: operators.coreos.com/v1
   kind: OperatorGroup
   metadata:
     name: cluster-logging
     namespace: openshift-logging 
   spec:
     targetNamespaces:
       - openshift-logging
   

1. Delete the configmap containing the Logging / Collection Dashboard

   oc delete cm  grafana-dashboard-cluster-logging -n openshift-config-managed

Actual results:

Verify that the Cluster Logging Operator is not watching the resource, then, it's not recreated. Similar to the reported in https://issues.redhat.com/browse/LOG-5308.

  $ oc get cm grafana-dashboard-cluster-logging -n openshift-config-managed
   Error from server (NotFound): configmaps "grafana-dashboard-cluster-logging" not found
  

Expected results:

The configmap `grafana-dashboard-cluster-logging` gets recreated by the ClusterLogging Operator as this last is watching the resource.

Workaround:

Follow the steps for "Upgrading the Red Hat OpenShift Logging Operator to watch all namespaces".