How would you handle a situation where a critical service you manage is experiencing significant downtime?