Every small team running Kubernetes has that upgrade. It has been on the board for three sprints. Everyone agrees it needs doing. Nobody wants to break production on a Thursday. Meanwhile, the cluster keeps getting further behind. If the same two…