A failure to ensure backup power for some of its computer servers led to the meltdown at Delta Air Lines this week that forced more than 2,000 flight cancellations through Wednesday afternoon, the airline’s top executive said.
Delta CEO Ed Bastian, in an interview with The Atlanta Journal-Constitution, said he takes “personal responsibility” for the episode and added new detail to its cause.
The problem, Bastian said, was that about 300 of the Atlanta-based airline’s 7,000 servers were not wired to backup power. Early Monday, when a power control module at Delta’s technology command center failed and caught fire, it caused a surge to a Georgia Power transformer and a loss of power.
It was initially assumed a Georgia Power outage started the chain of events, but Bastian said that’s not the case.
“Delta’s responsible for this,” said Bastian, adding, “The buck stops with me. I’m the CEO.”
Bastian said the backup power vulnerability “went undetected and should not have.”
The problem brought Delta’s operations to a temporary halt and caused a cascade of cancellations and delays that disrupted travel for hundreds of thousands of people around the world.
By late Wednesday the effects were receding but Delta still canceled more than 300 flights for the day.
Bastian, a longtime Delta executive who became CEO last spring, said only the government-ordered grounding after 9/11 caused a shutdown as complete as Monday morning’s after the computer systems failed.
Some flights resumed after a few hours but by then the airline had to deal with an avalanche of displaced customers and flight crews and out-of-position aircraft, and systems that came back up were operating slowly.
“When we first were there analyzing it in the early hours, all you saw was that there was a loss of power,” Bastian said. “We’re not here pointing fingers at Georgia Power.”
The servers that lacked backup power are a relatively small percentage of the total, but they are “meaningful in their impact,” Bastian said.
When other servers that were connected to dual power sources came back on, he said, “they did not get responses from the 300 servers… [which] caused the entire system to crash.”
Many reports questioned whether Delta and other airlines have let their technology grow old and vulnerable. Southwest Airlines recently suffered a similar snarl, and technology outages have hit other carriers as well.
Bastian acknowledged that “our infrastructure is dated, no question,” with “legacy mainframes.” But, he said, “I don’t think that was the problem.”
He said Delta spends about $1 billion a year on technology and this year hired a new chief information officer from insurance giant AIG and a new head of technology infrastructure from Marriott, who were “pulling together the next level of investment.”
“This is an area that we know was in need of investment. We have been investing in it,” Bastian said. “It’s not an area that we’ve been dismissive of.”
He also acknowledged: “I knew there was more to be done to make certain we didn’t have vulnerabilities such as this, and we’re in the process of continuing to monitor and upgrade.”
“The fact that this is something we were unaware of is unacceptable,” Bastian said.
Delta on Wednesday said some flights were still being delayed or canceled due to crews stuck in the wrong place or exceeding federal caps on on-duty hours.
Delta said it was focusing on moving flights through its Atlanta hub, the largest in its operation, and also used its Delta Private Jets subsidiary to get 40 customers from Atlanta to their destinations.
Many exhausted passengers were still suffering the effects of flight cancellations.
Mariah Paden spent close to 11 hours waiting in the Seattle airport, trying to get to Atlanta, after her flight was canceled.
“It was just really chaotic,” Paden said Wednesday. “There were a lot of people that were really upset…. I’m going to try to plan my flights differently just so I can get away from errors like this.”