When machineconfig fails to generate, we set upgradeable=false and degrade pools. The expectation is that the CO would also degrade after some time (normally 30 minutes) since master pool is degraded, but that doesn't seem to be happening. Based on our initial investigation, the event/degrade is happening but it seems to be being cleared.

    4.18

    Should be always

    1. Apply a wrong config, such as a bad image.config object:
spec:
  registrySources:
    allowedRegistries:
    - test.reg
    blockedRegistries:
    - blocked.reg
    
    2. upgrade the cluster or roll out a new MCO pod
    3. observe that pools are degraded but the CO isn't