Amazon.com announced that its cloud services returned to normal on Monday afternoon following an internet outage that disrupted thousands of sites globally, including major apps like Snapchat and Reddit.
However, Amazon noted that some AWS services still had a backlog of messages that would take a few hours to clear.
AWS, which hosts applications and processes for companies worldwide, experienced a disruption that affected workers from London to Tokyo. This prevented many from performing regular tasks such as paying for services or changing travel plans. Users reported ongoing issues with services like Venmo and Zoom on Monday afternoon.
Stay informed with the latest updates. Follow KT on WhatsApp Channels.
This incident marked the largest internet disruption since last year’s CrowdStrike outage, which impacted technology systems in hospitals, banks, and airports, highlighting the vulnerabilities within global technology networks.
This was at least the third significant internet disruption in five years involving AWS’s northern Virginia data center, known as US-EAST-1.
Amazon did not provide further details on why this data center is frequently affected. The issues were linked to the Domain Name System (DNS), which hindered applications from locating the correct address for AWS’s DynamoDB API, a crucial cloud database for storing user information.
Network Health Monitor Identified as Root Cause
Earlier, AWS attributed the outage to a subsystem responsible for monitoring the health of network load balancers used for traffic distribution across servers.
The problem originated within the “EC2 internal network,” part of Amazon’s “Elastic Compute Cloud” service offering on-demand cloud capacity.
Amazon stated that by 3pm PT (10pm GMT), all AWS services resumed normal operations, although services like AWS Config, Redshift, and Connect were still processing message backlogs.
Ken Birman, a Cornell University computer science professor, emphasized the need for better fault tolerance in software development. He noted that AWS provides tools for developers to safeguard against data center issues, and developers can also back up data with other cloud providers.
“Companies that cut costs and skip important steps to quickly launch applications without adequate protection against outages are the ones that should be examined closely,” Birman told Reuters.
AWS Site Known for Previous Outages
AWS, the world’s largest cloud provider, offers computing power, data storage, and digital services to companies, governments, and individuals, followed by Microsoft’s Azure and Google’s Cloud.
Server disruptions can lead to outages across various websites and platforms, impacting apps for food delivery, gaming, and airline systems that depend on its infrastructure.
AWS’s status page indicated Monday’s outage began at its US-EAST-1 location, its oldest and largest web services site, which also experienced outages in 2021 and 2020.
Documentation on AWS’s website states that US-EAST-1 is often the default region for many AWS services.
‘Fragile Infrastructures’
The issue highlights the interconnectedness of digital services and their dependence on a few global cloud providers, where a single glitch can disrupt business and everyday activities, experts said.
“This outage underscores our reliance on relatively fragile infrastructures,” stated Jake Moore, global cybersecurity advisor at ESET.
In the UK, Lloyds Bank, Bank of Scotland, and telecom providers Vodafone and BT were affected, as was the UK tax authority’s website, according to Downdetector’s UK site.
“The core issue is the reliance of these major companies on a single service,” said Nishanth Sastry, director of research at the University of Surrey’s Computer Science Department.
Ookla, owner of Downdetector, reported over 4 million user issues due to the incident.
“For large businesses, hours of cloud downtime mean millions in lost productivity and revenue,” said Ryan Griffin, US cyber practice leader at McGill and Partners.
Despite the disruption, Wall Street remained steady, with Amazon shares rising 1.6% to $216.48.
Outage Impacts Apps from Snapchat to Venmo
Ookla reported that the outage affected at least a thousand companies, including apps like Reddit, Roblox, Snapchat, and Duolingo.
Companies like AI startup Perplexity, cryptocurrency exchange Coinbase, and trading app Robinhood all reported disruptions linked to AWS.
Amazon’s services, including its shopping site, Prime Video, and Alexa, were also disrupted.
Gaming platforms like Fortnite, Clash Royale, and Clash of Clans were impacted, as was Lyft in the US.
Signal President Meredith Whittaker confirmed on X that the messaging app was affected by the outage, although Elon Musk stated his platform continued to function.
