Global Web Thrown Into Chaos by 1/2 Day Amazon Cloud Outage

The worldwide web was thrown into disarray on Monday, 20th October, following a massive, day-long outage of Amazon Web Services (AWS), which severely disrupted a significant proportion of online services globally and exposed the alarming dependency of the world on US tech giants.

Amazon Web Services (AWS), the world’s leading cloud platform, confirmed it had resolved the protracted failure after nearly 15 hours of crisis management.

The firm, a subsidiary of Amazon, announced just before 10pm GMT that the problem was completely fixed, though it anticipated “a total restoration… within the next two hours” as a backlog was processed.

The localised failure began shortly after 7am GMT in historic AWS data centres near Washington, causing widespread chaos.

Banks, online games such as Fortnite and Roblox, streaming services including Disney+ and Prime Video, and everyday applications like Airbnb, Zoom, and Snapchat all experienced disruptions or total interruptions.

The outage, which blocked payments, halted deliveries, and prevented other professional and private activities, starkly illustrated the global dependency on US tech infrastructure.

AWS, which provides shared data centres, private servers, and AI tools, is the world’s largest “cloud” computing platform, commanding nearly a third of the rapidly expanding global market.

Its closest competitors are fellow US firms, Microsoft Azure and Google Cloud, which share the second third of the market, according to Synergy Research Group.

Vulnerability Exposed

British financial analyst Michael Hewson pointed out that the incident raises “serious questions” about the wisdom of companies “outsourcing all or part of their essential infrastructure to a small group of third-party vendors” simply to save money on hosting.

Junade Ali, a cybersecurity expert at the UK’s Institution of Engineering and Technology (IET), warned the outage “highlights the challenges associated with dependence” on foreign-based service providers like Amazon, Microsoft, and Alphabet (Google).

Analyst Gadjo Sevilla of Emarketer emphasised that the “excessive reliance on a single provider now threatens more than just service availability: it jeopardises brand reputation and customer trust,” suggesting AWS clients need costly redundancy strategies.

AWS initially suspected a problem with the Domain Name System (DNS), which guides online requests. However, after a brief respite, difficulties resumed around 3pm GMT, affecting US services such as the game Battlefield, Delta Air Lines, and the popular payment service Venmo.

AWS later identified a more serious “root cause”: “an internal subsystem responsible for monitoring the proper functioning of network load balancers”—the central control tower of the system, not just the navigation.

Despite AWS dividing the world into forty “regions,” each with three distinct structures designed to mitigate total network failure, 20th October’s incident proved that fundamental requests continue to pass through the older and most significant AWS centre, US-East-1, located in northern Virginia.

Source: Defi Media