In order to provide better identification of which operator(s) are at or near the root cause of failures we want to replace the existing operatorconditions logic with a smarter framework that can evaluate the operators by order of dependency.

1. Select some key operators and figure out the dependencies between them, get an ordered list, check them one by one, once any abnormal operators are found, only generate one failed case in the junit xml, other operators would be marked as “skipped”.
   For the left operators (or even new operators introduced in the future), if the above core operators' health check gets passed, run a simple check just like before, not care about dependency (if find out some new dependency, prompt it to the above key operator ordered list). if the above core operators' health check failed, all the left operators would be marked as “skipped”.

The key operators' order maybe look like:
etcd -> network -> cloud-credential -> dns -> kube-apiserver -> kube-controller-manager -> openshift-apiserver -> openshift-controller-manager -> machine-api -> machine-config -> ingress -> storage -> image-registry -> authentication -> console

1. Develop the above logic with shell script to replace the line
2. Run it for days to validate if it meets the requirements
3. If needed, convert it to a golang tool
   1. We need to discuss the need for a golang tool and if it requires a new step
   2. We would like to be able to extend this framework as we go and that implies golang