NetApp has said none of its technology was involved in the Australian Virgin Blue airline reservation system crash on Sunday.
On Sunday and Monday a 21-hour outage of the New Skies system, hosted at Navitaire's Sydney data centre, on Monday caused mass delays to Virgin Blue's customers as the airline reverted to a manual system. The airline had to pay accommodation costs for some customers and the total cost of the outage could run into millions of Australian dollars.
SearchStorage ANZ reported that NetApp said: "no NetApp technology nor any technology acquired from NetApp was linked to the recent outage at Navitaire.”
Navitaire is a business processing outsourcing subsidiary of Accenture and offers New Skies as a private cloud service to a large number of airlines. The crash started when there was a failure in Navitaire's solid state disk server infrastructure. The company uses a Texas Memory Systems RamSan 500 NAND solid state memory array, whose storage resources are presented to users via a NetApp V-Series controller front end. The V-Series was not involved in the outage and nor was any other NetApp kit, according to NetApp's statement.
It is unusual for a solid state drive array to have a failure that causes software using it to crash or that corrupts data in the array to such an extent that a complete airline reservation system, whose database is resident in the solid state device, crashes.
The RamSan-500 offers up to 2TB of cached Flash RAID and 64 GB of DDR cache in a package that can produce 100,000 random sustained IOPS. The data in RAM is protected with ECC and Chipkill. The RAM cache is battery-backed up with redundant batteries. The flash storage is deployed in nine RAID-3 protected, hot swappable, modules. Each module also has ECC memory layouts, wear-leveling, and bad-block retirement.
With this set of protection features it is hard to understand how a RamSan hardware failure could cause the outage. Texas Memory Systems has said it has been working non-stop with Navitaire on the issue to find its root cause.
Navitaire's statement on the outage to Virgin Blue said it tried to "repair the device" and this attempt delayed the "cutover to a contingency hardware platform." Why that cutover wasn't managed for 21 hours is a mystery.
Navitaire has a service-level agreement with Virgin Blue requiring to it to fix computer system failures in a short time. In this outage it did not do so, suggesting that its disaster recovery and failover procedures were inadequate to the demand put on them
Richard Branson, the head of the Virgin Group, has apologised to affected customers for the computer systems failure.
The New Skies system failed again this morning, between 5.10am and 7am Australian Eastern Standard Time. Virgin Blue switched to a manual process again. Virgin Blue said in a statement that it: "experienced a slowdown of its network at the start of the day which caused some consequential delays."
Virgin Blue selected Navitaire's New Skies platform in June, 2009. Navitaire says "New Skies is a comprehensive airline passenger sales and management solution providing capabilities for integrated Internet booking, call centre reservations, travel agency global distribution connectivity, inter-airline and alliance code-share itineraries, real-time reporting, ancillary revenue generation and departure control."
There have been problems reported previously with the New Skies software, with Virgin Blue, Jetstar and RyanAir all encountering system failures.
How Accenture (motto: "Go on; be a tiger") could design and implement an airline reservation system that crashes for 21 hours with no fast fail-over or business continuance facility is puzzling. The company would obviously design in such a facility for vital software operating in real-time with thousands of customers a day, wouldn't it?
Instead of being a tiger it put customer Virgin Blue in a hole and kept it there for nearly a whole day. ®