Context
This came up during #7688.
What we need to do
We should automatically run a hub health check after a server startup failure alert. If the check succeeds, we should resolve the alert, if it doesn't maybe we can find a way to leave a note in the incident about it.
Definition of Done