A significant outage in Amazon’s cloud computing network on Tuesday badly affected services at a wide variety of U.S. corporations for more than five hours, highlighting how concentrated the industry of keeping the internet operating has become.
The Amazon Web Services disaster mostly hit the eastern United States, but it also disrupted everything from airline bookings and auto dealerships to payment applications and video streaming services, as well as Amazon’s own large e-commerce business. This included The Associated Press, whose publishing system was down for most of the day, severely hampering its capacity to post news reports.
Amazon has yet to say anything about what went wrong. In fact, Amazon’s communications on Tuesday were limited to terse technical explanations on an AWS dashboard and a brief statement delivered via spokesperson Richard Rocha that acknowledged the outage had impacted Amazon’s own warehouse and delivery operation but said the company was “working to resolve the issue as quickly as possible.”
Approximately five hours after multiple firms and other groups began reporting troubles, the company announced in a statement on the AWS status page that it had “mitigated” the underlying problem that was causing the outage, which it did not specify. Some impacted businesses needed many hours to properly inspect their systems and restart their own services.
Amazon Web Services was formerly led by Amazon CEO Andy Jassy, who took over for founder Jeff Bezos in July. Amazon’s cloud-services business is a big profit source for the company. According to Synergy Research, it controls about a third of the $152 billion cloud services industry – a higher stake than its nearest competitors, Microsoft and Google combined.
According to Carl Malamud, a technologist and public data access campaigner, the AWS outage demonstrates how much Big Tech has twisted the internet, which was initially meant as a distributed and decentralized network capable of surviving mass tragedies such as a nuclear assault.
“We’re breaching that fundamental notion when we put everything in one place, whether it’s Amazon’s cloud or Facebook’s monolith,” said Malamud, who created the internet’s first radio station and subsequently placed a critical US Securities and Exchange Commission database online. “We saw that when Facebook became the tool of a big disinformation effort, and we just saw it today with Amazon’s collapse.”
Widespread and typically protracted outages caused by single-point failures are becoming more prevalent. In June, the content distributor behind-the-scenes Fastly had a malfunction that momentarily brought down dozens of prominent internet sites, including CNN, The New York Times, and the government home page of the United Kingdom.
Then, in October, Facebook — now known as Meta Platforms — blamed a “erroneous configuration update” for an hours-long global outage that knocked out Instagram and WhatsApp as well as its own platform.
This time, issues began about midday on the United States’ East Coast, according to Doug Madory, director of internet monitoring at Kentik Inc, a network intelligence provider. One of the most renowned names hit was Netflix, which suffered a 26 percent decline in traffic to its streaming service, according to Kentik.
Customers attempting to book or change flights with Delta Air Lines were unable to connect to the carrier. “Delta is working hard to restore functionality to its AWS-supported phone lines,” stated Delta spokesman Morgan Durrant. Customers were urged to utilize the airline’s website or mobile app instead, and the company apologized.
Southwest Airlines, located in Dallas, said it shifted to West Coast servers after several airport-based systems were disrupted by the outage. Customers were still reporting problems to DownDetector, a popular user outage reporting clearinghouse, more than three hours after they began. According to Southwest spokesperson Brian Parrish, there were no major flight problems.
According to Toyota spokesman Scott Vazin, the company’s U.S. East Region for dealer services has gone down. Apps for accessing inventory data, monthly payment calculators, service bulletins, and other stuff are available from the firm. Over 20 applications were impacted.
People attempting to use Instacart, Venmo, Kindle, Roku, and Disney+ also experienced troubles, according to DownDetector. The McDonald’s app was also unavailable. However, American, United, Alaska, and JetBlue were untouched.
Madory stated that he found no cause to suspect any wrongdoing. According to him, the recent cluster of big outages demonstrates how sophisticated the networking sector has gotten. “More and more of these outages are the result of automation and centralization of management,” he explained. “This results in disruptions that are difficult to totally avoid owing to operational complexity but have a significant impact when they occur.”
It was unclear whether or not the outage affected the federal government. In an email response to queries, the US Cybersecurity and Infrastructure Security Agency stated that it was working with Amazon “to assess any possible consequences this outage may have for government agencies or other partners.”