> As you probably know when manual maintenance is being done you might > take advantage of that knowledge by e.g. unmanaging the cluster during that time. We have a LOT of clusters to manage for a small team. I would prefer to just have them survive brief network outages that last for a known expected duration. --Eric