Imagine managing a gitlab instances. You're doing it for years and have a stable environment. If one of the instance gets hammered by external traffic or one of your data center dies, will you call the gitlab devs to deal with it ?