There are some times when a maintenance is potentially super disrupting such as when node replacements occur due to AZ rebalancing. It would be nice if we can set the maintenance period per cluster vs per project to be able stagger maintenance periods. so that not every cluster goes under such disruptive maintenance.