In version 4.20.8, a default CronJob named machine-config-nodes-crd-cleanup exists in the openshift-machine-config-operator namespace.
Currently, this CronJob uses the following nodeSelector:
```
node-role.kubernetes.io/master
```

For a more future-proof and standardized configuration, this should be updated to:
```
node-role.kubernetes.io/control-plane
```

Using control-plane aligns with current Kubernetes and OpenShift naming conventions and ensures compatibility with future releases. This change is also recommended in the referenced KCS article for better role naming and long-term support.

https://access.redhat.com/solutions/7002512


Additionally, when this CronJob is manually executed from the OpenShift web console, a popup warning message is displayed. This behavior is related to the usage of the deprecated master node role and further highlights the need to update the nodeSelector to control-plane.

~~~
Admission Webhook Warning

CronJob machine-config-nodes-crd-cleanup violates policy 299 - "spec.jobTemplate.spec.template.spec.nodeSelector[node-role.kubernetes.io/master]: use \"node-role.kubernetes.io/control-plane\" instead"
~~~

    100%

    1. Run the machine-config-nodes-crd-cleanup cronjob manually from the web console
    

   Warning: spec.jobTemplate.spec.template.spec.nodeSelector[node-role.kubernetes.io/master]: use "node-role.kubernetes.io/control-plane" instead
 

   nodeSelector parameter should be node-role.kubernetes.io/control-plane in the cronjob.