The Cluster Version Operator should improve its wording in Failing=True messages. For example, the following message could be interpreted as such that the upgrade has failed, and the CVO was unable to apply the upgrade:

```
Unable to apply 4.18.20: wait has exceeded 40 minutes for these operators: kube-apiserver
```

In reality, the upgrade does not fail and is rather failing. The CVO is continuously trying to apply the update and will never stop. The meaning of the message is to convey to the cluster administrator that the CVO is having unexpected difficulties. The CVO was waiting for the kube-apiserver operator to update for 40 minutes, which should be plenty; however, the operator still has not updated, and we recommend an investigation by the admin. However, the CVO is still trying to apply the update.

This is not very clear from the message.

    1.
    2.
    3.
    

   An already discussed solution could be to add a prefixed message at the beginning of all Failing=True messages, where we explain that the CVO will never stop trying to reconcile and where we state that the CVO has unexpected difficulties, and we recommend an admin investigation.