Description of problem: Customer found an issue with creating a scheduled job that appears to break their OpenShift cluster as a whole. They first noted pods not being able to pull images from the docker registry. It was also reported that certain pods that use the internal network were unable to do so. We also saw certain oc commands hang and others cause the terminal to freeze entirely. We are able to restore everything to functional status by restarting the atomic-openshift-master-{api,controllers} services As such we attempted to increase the log levels on the atomic-openshift-master-{api,controllers} and collected the logs I will be attaching to the bug shortly. Version-Release number of selected component (if applicable): openshift 3.4.1.2