1. Proposed title of this feature request
Capture MachineConfigDaemon Events in the Operator events
2. What is the nature and description of the request?
Everytime the config on the node is changed, the MCD observes this error:
~~~
unexpected on-disk state validating against rendered-worke
~~~
3. Why does the customer need this? (List the business requirements here)
This issue was observed on several cases and this is causing issue as we have to check through the MCD logs for the misconfigured or differing configs.

4. List any affected packages or components.

• MCO

ISSUE explanation and requirement:

• whenever any change is made to the config on node which is managed by MCO, the mcd on that node crashes with very few logs and upon investigation we get to know that some file is missing or is having mis-matching config parameters
• When customer does any modification to the config file on the node and as the modified file conflicts witht the configuration provided by machine-config, the daemon on that node crashes with the error: "unexpected on-disk state validating against rendered-worke", we have to go through the pod logs and see which file has conflicting changes and then upon fixing the changes the MCD restarts which then allow the node upgrade proceeds.

Suggestion:

• If the Operator checks the MCD logs on the node and captures the file conflicts and points which files are modified and giving "unexpected on-disk state validating against rendered-worke", we can directly check the output of `oc get co machine-config` anc check why the operator is degraded before applying the upgrade as this type of issue is only noticed during upgrade as few nodes do not get upgraded.