PARIS:
A massive outage in Amazon Web Services (AWS) on Monday crippled internet operations across the world, knocking out thousands of websites, mobile apps, and even some banking services for several hours before engineers began restoring connectivity.
The disruption, one of the largest since last year’s CrowdStrike malfunction, disrupted platforms including Snapchat, Reddit, Airbnb, Disney+, Prime Video, Perplexity AI, Duolingo, Signal, and Fortnite.
Financial platforms such as PayPal’s Venmo, Coinbase, and Robinhood were also affected, along with British banks including Lloyd’s and Bank of Scotland. Government websites in the UK, including HMRC, faced interruptions, while telecom networks Vodafone and BT reported degraded services.
Outage tracker Downdetector reported over 9,000 user complaints at the peak, as AWS engineers rushed to fix what they later identified as a fault in the “EC2 internal network”, linked to a malfunction in the subsystem that monitors the health of Amazon’s network load balancers.
The issue originated at AWS’s US-East-1 site in northern Virginia – its largest and oldest data hub, which has previously suffered outages in 2020 and 2021. AWS said the root cause was an internal monitoring component failure affecting Domain Name System (DNS) operations, which route internet traffic to the correct destinations.
The fault triggered widespread connection timeouts across its cloud-based services. The company said the problem was “fully mitigated” after several hours, though lingering errors slowed the return to normal operations.
Ookla, which operates Downdetector, said more than four million users worldwide faced disruptions. Amazon confirmed its own services – including its e-commerce platform, Prime Video, and Alexa – were among those hit. By late evening, most applications had gradually stabilised, but AWS continued reporting elevated error rates in some regions.
Cybersecurity experts said the outage exposed the fragility of the world’s digital infrastructure, which relies heavily on a handful of “hyperscalers” – Amazon, Microsoft, and Google – that dominate the global cloud computing market.
“This incident highlights the dependency we have on relatively fragile infrastructures,” said Jake Moore, Global Cybersecurity Advisor at ESET. University of Surrey researcher Nishanth Sastry warned that concentrating digital services under a few providers made the entire system “vulnerable to cascading failures.”
Analysts drew parallels to the July 2024 global outage caused by a faulty software update from cybersecurity firm CrowdStrike, which crippled systems in hospitals, airports, and banks. The AWS breakdown, though unrelated, again demonstrated how technical issues in one region can ripple across global networks.
Amazon Web Services currently controls about 30% of the world’s cloud computing market, followed by Microsoft’s Azure at 20% and Google Cloud at 13%, according to Synergy Research Group.
These platforms provide essential computing power and data storage for millions of businesses and consumers – from email systems and financial transactions to gaming and streaming.
As systems recovered late on Monday, experts said the episode would likely reignite debate over the world’s growing dependence on a few tech giants to keep the internet running.
“A single point of failure in one data centre can bring parts of the global economy to a halt,” said cyber analyst Rimesh Patel. “It’s a sobering reminder of how deeply interconnected our digital lives have become.”
