While looking into bug 1647514 we discovered that when some upgrade task fails on one controller, the upgrade stops on that controller, but further tasks are applied on the remaining controllers. I think this is generally not desirable, we should stop the upgrade fully as soon as we hit an error.
Master merged, backport to rocky proposed.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2019:0878