THAILAND – A widespread outage at Amazon Web Services (AWS) knocked dozens of popular websites and apps offline for several hours on Wednesday, leaving millions of users and businesses unable to reach services that depend on the cloud giant.
Mass disruption across apps and sites
The failure began in the early morning and affected a mix of household names and niche platforms. Users reported problems accessing social and gaming services such as Snapchat, Reddit and Roblox, streaming and broadcast platform Max, and financial apps including Venmo and Coinbase. Even Amazon’s own offerings — its shopping site, Prime Video and the Alexa voice assistant — experienced interruptions.
The disruption lasted roughly three hours in many regions. For businesses and consumers that rely on near-instant access to online services, the outage was an unusually long interruption. Contact centres and customer-support lines saw heavy volumes as organisations scrambled to communicate with users.
Amazon pointed to DNS and database issues
Amazon said preliminary investigations indicated the problem originated in the Domain Name System (DNS), which translates website names into the numeric IP addresses computers use to route traffic. When DNS services fail, the internet’s normal name-resolution process breaks down and users cannot reach the intended sites.
A company spokesperson said, “We have observed clear signs of recovery,” and added that engineers believed a networking error that affected database services was to blame. AWS highlighted DynamoDB — a key cloud database used by many customers to store application data — as particularly impacted, noting that access to stored customer data had been unavailable for several hours.
Businesses felt the immediate impact
Companies that had built their services on AWS infrastructure faced a range of problems. Some platforms were forced to suspend operations entirely; others operated in a degraded state. Organisations that depend on real-time data feeds and authentication services reported the most acute effects.
The uncertainty over when systems would be fully functional added to the strain, with teams working to reroute traffic and apply temporary fixes while waiting for AWS to restore full service. Amazon urged customers to clear local caches after services came back online to speed recovery of normal performance.
Experts warn about concentration risks
Cybersecurity and cloud experts said the outage highlighted the risks of heavy concentration of internet infrastructure in the hands of a few providers. “It shows how risky it is when the world relies on just a few large cloud providers,” said Mike Chapple of the University of Notre Dame. “If a central service goes down, a significant portion of the internet can suddenly go quiet.”
The migration of data and applications to major cloud platforms has accelerated because of cost savings and operational ease. That convenience, experts say, can come at the cost of resilience when a provider experiences urgent technical failures.
Recovery and aftermath
By about 6:35 a.m. US time, Amazon reported that systems had been restored for most customers. While Amazon said no customer data had been lost, some services continued to take additional hours before all features returned to normal. The company said it would produce a full incident report in the coming days after completing its internal review.
Wider implications and next steps
The outage echoed similar incidents in recent years — including a disruption last year linked to a security vendor that affected hospitals, banks and airports — underscoring the fragility of interconnected systems. For many businesses, the episode will prompt renewed discussions about redundancy, multi-cloud strategies and contingency planning.
Amazon said it was continuing forensic work to determine precisely what triggered the DNS and database failures. In the meantime, companies that rely on cloud services will be reviewing their preparedness for the next unexpected disruption.