In early 2000, Northwest Airlines suffered a network outage at its' worldwide IT operations center, in Eagan MN. This data center houses all airline flight operations and flight safety systems. The network failure disrupted all IT operations in the data center, ultimately causing the grounding of all NWA aircraft worldwide for over 24 hours. Thousands of flights were canceled, hundreds of thousands of passengers were inconvenienced and millions of dollars were lost. Normal flight schedules and flight operations were disrupted for days.

The outage was caused by cable cut. A trenching machine shredded a major multi-pair fiber optic feed and multi-pair copper cable feed that fed the airline's hardened data center. The cut cable was owned by the regional carrier at the time, Qwest Communications. It was part of fully redundant fiber optic ring and service called Self Healing Network Service, or SHNs. By design, if network service is disrupted by a cut cable, SHNs is guaranteed to automatically reroute the traffic over a diversely routed fiber optic facility. The secondary ring, as specified in the tariff, was supposed to be no closer than 35 feet to the primary ring at all points; clearly it was not. As Murphy's Law has taught us countless times, 'what can go wrong will go wrong'. The primary and secondary fiber-optic cables were located in the same conduit. This violated the specifications documented in the tariff and guarantees provided by the carrier. A single digging activity cut both SHNs primary and secondary cables and crippled operations at the airline's data center.

It gets more a little more complicated.

• The SHNs cables were severed by McLeod Communications, an independent Qwest competitor at that time. They were installing a new fiber-optic facility to serve the area (to compete with Qwest).
• The cable was actually cut by a McLeod Communications installation subcontractor.
• The root cause of the problem was eventually traced to defective network facility planning software used and operated by Qwest. This complex facility planning software application directs network planners in the physical placement of their fiber-optic facilities. A defect in the code directed Qwest network facility installers to bury the primary and secondary cable path in the same conduit

A lot of money and customers were lost and lawsuits remained unsettled for years, because a well-intentioned physical layer network recovery mechanism failed.

Who do you feel is at fault and responsible for the network outage? Justify and support your opinion.