@Gargron Yeah, we were scaling up and down with intentional high roll over events(Daily rebuilds of all nodes). If we were more stable I would have loved it. Amazingly easy to reason about vs K8s.