This is an interesting & relevant question. What are the best practices in managing large "dev" clusters that typically have multiple users (developers?). How a

This is an interesting & relevant question. What are the best practices in managing large "dev" clusters that typically have multiple users (developers?). How are "infra failure" tests performed? (or are these clusters not the right platform for such testing)