Amazon Cloud Outage Disrupts Services Worldwide, Now Resolved

A significant outage of Amazon’s cloud computing service, Amazon Web Services (AWS), disrupted internet access globally on March 11, 2024. The issue, which began early in the morning, affected a wide array of online services including social media, gaming, food delivery, streaming, and financial platforms. By Monday evening, Amazon confirmed that the outage had been resolved, restoring normal operations.

The disruption highlighted society’s increasing reliance on a few key technology providers. As services began to recover three hours into the outage, AWS reported progress on its health dashboard. By approximately 18:00 Eastern Time, the company stated that “services returned to normal operations.” AWS underpins many of the internet’s critical infrastructures, serving major organizations across various sectors.

Cybersecurity expert Mike Chapple, a professor at the University of Notre Dame, described the recovery process as “slow and bumpy,” noting that smaller interruptions might occur as fixes are implemented. Chapple compared it to a large-scale power restoration, where some areas may experience intermittent issues until repairs are fully completed.

Amazon attributed the outage to problems with its domain name system, which is essential for converting web addresses into IP addresses. This malfunction hampered the ability of numerous websites and applications to function properly. According to DownDetector, a service that tracks online outages, over 11 million user reports of problems were logged across more than 2,500 companies.

Among the affected services were popular platforms such as Snapchat, Roblox, Fortnite, the online broker Robinhood, and the McDonald’s app. Streaming services like Netflix and Disney+ also faced disruptions. Furthermore, Amazon’s own offerings were impacted, with users unable to access Ring doorbell cameras or utilize Alexa-powered devices.

The outage particularly affected education, with many college and K-12 students unable to access homework or course materials. Damien P. Williams, a professor of philosophy and data science at the University of North Carolina at Charlotte, expressed frustration over the inability to grade assignments. His difficulties were echoed by students from various institutions, as the widely used educational platform Canvas was rendered inoperable.

Canvas reported that it serves approximately 50% of college and university students in North America, including all Ivy League schools. Ohio State University informed its 70,000 students that online course materials may be inaccessible and encouraged them to connect with instructors for alternative arrangements.

The first signs of trouble appeared at around 03:11 Eastern Time on March 11, when AWS acknowledged increased error rates for multiple services in the US-EAST-1 Region. As the day progressed, the company reported “significant error rates” and assured users that engineers were actively working on the issue. By 06:00, AWS noted that recovery was underway for most affected services, although full resolution took several more hours.

This incident is not the first time AWS has faced significant outages. In late 2021, a similar disruption affected numerous companies for over five hours. Previous outages also occurred in 2020 and 2017, underscoring the vulnerability of the internet’s infrastructure due to reliance on a small number of service providers.

Patrick Burgess, a cybersecurity expert from BCS, The Chartered Institute for IT, commented on the far-reaching impact of such outages. “The world now runs on the cloud,” he stated. “When there’s an issue, it can be really impactful across many online services.” Burgess emphasized that these kinds of outages are typically resolved relatively quickly and indicated that this incident did not appear to be the result of a cyberattack.

Despite the frustrations caused by the outage, experts generally agree that established protocols exist to handle such incidents, and most outages are typically resolved within hours rather than days. The rapid recovery from this latest disruption serves as a reminder of the critical role that cloud services play in daily life.