All instances are up and running again.
Due to a host failure, about 300 instances are rebooting. Takes up to 20 mins until all are online again.