CloudBees Replication Parent 1768

If the Jenkins startup process was terminated by a boot failure (a fatal error, such as incorrect CasC bundle contents) in a managed controller, the pod would not be ready or not attempt to restart the container; it would just hang until manually deleted, even if the underlying issue had since been corrected. For a High Availability (HA) controller, this could mean that a rolling restart would not make progress. Now, the container exits and the pod will be shown as CrashLoopBackOff status and periodically try again to start the controller process. If the condition causing the error is corrected, the next container restart should allow the pod to become ready, and the controller restart to complete.