The relay node had been offline for eleven minutes before anyone noticed. By then three downstream systems had made decisions based on stale data, and two of those decisions had already been executed. The third was still pending, which was the only reason the audit log would later show what had happened at all. The engineer who wrote the incident report used the phrase 'cascading assumption failure.' Her manager changed it to 'data synchronization issue.' Her manager's manager removed the section entirely and noted that all systems were now operating normally.