Looks like when the new reboot plan that handles rebooting nodes and waiting is using the last boot time to check if a node rebooted. Unfortunately the precision is not very high which means if the last boot time before the reboot was 3:10pm and then the reboot occurs and happens very quickly ( within the same minute ) and the last boot time obtained after the reboot is 3:10pm then the plan execution is stuck in an infinite loop forever.
Time Stamp comparion that was happening was:
12/11/2018 3:11 PM ( before reboot )
12/11/2018 3:11 PM ( after reboot )
To resolve this milliseconds needs to be taken into account or some other means of getting finer grained precision.