Amazon Restores Services Following Worldwide Outage Interruptions Major Websites, Banks, and Snapchat

Amazon claims that its cloud services are back up and running following a broad outage that caused disruptions to international platforms like Reddit, Snapchat, and Lloyds.

AWS reports that it has fixed a significant outage that prevented users from accessing some of the biggest websites and applications in the world for a considerable portion of Monday.

Over 1,000 online platforms, such as Reddit, Snapchat, Lloyds Bank, and Halifax, experienced issues that were linked to the heart of Amazon’s cloud computing business in the United States.

Over 11 million users worldwide reported technical problems during the outage, according to Downdetector, a website outage tracker.

Amazon verified that all AWS services had “returned to normal operations” at approximately 23:00 BST, after the throttling of some system components to resolve the problem.

“This episode has brought to light the interdependence of our infrastructure,” said Alan Woodward, a professor at the University of Surrey. This indicates that issues might arise with even the biggest third-party providers, as many internet services depend on them for their physical infrastructure. Small mistakes—which are frequently human—can have a big, lasting effect.

The outage, which started at around 7:00 BST, affected a variety of websites and applications, including the language-learning website Duolingo and online games like Fortnite. Over four million problem reports were submitted in a matter of hours by Downdetector, which is more than twice as many as would be expected on a usual weekday.

Professor of computer technology at Notre Dame University Mike Chapple likened the restoration attempts to a power outage. It is similar to a widespread power outage. According to him, crews begin working to try to get it back online. “The power may flicker a few times,” he said, adding that Amazon might have at first “only addressed the symptoms” rather than the root reason.

Although the cause of the outage has not yet been fully disclosed by Amazon, the company stated that it “appears to be related to DNS resolution of the DynamoDB API endpoint in US-EAST-1.” By converting website names into computer-readable numbers, the Domain Name System, or DNS, acts as the internet’s equivalent of a phone book.

The event highlighted the dangers of cloud dependency, according to Cloudflare CEO Matthew Prince. “Everybody has bad days, and Amazon had a bad day today,” he said to the BBC. “Cloud computing has many wonderful features, including the ability to scale. However, if there is a disruption like this, many of the services we depend on may be unavailable.

As the director of the Future of Technology Institute, Cori Crider compared the disruption to “a bridge collapsing.” “A vital component of the economy has collapsed,” she added, referring to the dependence on Amazon, Microsoft, and Google, which control around 70% of cloud computing, as “unsustainable.”

“A significant portion of the economy is lost when something like this collapses, once you have a concentrated supply in a small number of monopoly providers,” Crider said. “Rather than depending on a few American monopoly platforms, we should actually consider attempting to purchase additional local services. Our economy, security, and sovereignty are all at danger, thus in order to strengthen our markets against such shocks, we must consider structural divisions.

According to Ken Birman, a professor of computer science at Cornell University, businesses that use AWS are also accountable. “Businesses that use Amazon haven’t been careful enough to incorporate security measures into their apps,” he claimed.

He went on to say that developers ought to put more money into protecting important systems because “we know how to make these systems stronger, and we know how to do it securely.”

Legal repercussions could yet result from the fallout. Following a comparable failure involving CrowdStrike last year, Delta Airlines has been suing to recoup costs of more than $500 million. The airline said that it had to manually reset 40,000 servers even after the problem was resolved, which resulted in significant flight delays for a number of days.