How do you manage root-cause analysis, containment, and corrective/preventive actions when an issue hits operations?