A technical failure in Amazon Web Services (AWS) disrupted the Alexa virtual assistant and various digital services in the early hours of October 20, 2025, affecting users in 190 countries, including Brazil. The issue, identified at 3 a.m. Brasília time, impacted Echo devices, streaming platforms, and financial apps. The cause was a Domain Name System (DNS) error in the US-EAST-1 region in the United States. AWS restored 80% of services by 8 a.m., though some residual delays persisted.
The outage caused slowdowns in services like Prime Video and Amazon’s shopping site. External platforms, such as Snapchat and Fortnite, also experienced disruptions. In Brazil, Downdetector recorded a 40% surge in complaints above the daily average, peaking between 4 a.m. and 6 a.m. Users reported difficulties connecting smart devices, like lights and thermostats, to the internet.
- Complaints in Brazil rose 40% on Downdetector.
- Prime Video and online games faced interruptions.
- Financial services experienced transaction delays.
Technical origin of the outage
The failure stemmed from the DynamoDB API endpoint, an AWS database service. The DNS issue affected authentications and data transfers, spreading to the Elastic Compute Cloud (EC2). The US-EAST-1 region, handling 30% of AWS’s global operations, amplified the impact.
AWS engineers detected high error rates and applied emergency fixes. Monitoring CloudTrail logs helped identify residual issues, reducing risks of further disruptions.
Impacts on digital services
The outage affected entertainment platforms, with Prime Video pausing streams during peak hours. Online games like Roblox and Clash Royale reported mass disconnections. Social media platforms, such as Reddit, saw 50% traffic drops.
Financial apps like Venmo and Coinbase faced transaction blocks. In Brazil, delivery services like iFood reported tracking failures. Universities noted unavailability in online learning portals.
In retail, Amazon’s website processed orders slowly, affecting 14,000 users, per Downdetector. The outage highlighted reliance on a single cloud provider.
AWS mitigation actions
AWS issued status updates every 30 minutes, detailing progress. Alternative traffic routes were implemented to bypass the faulty DNS.
Teams prioritized critical operations, processing queued requests in Lambda, a serverless computing service. Failover tests in secondary regions, like US-WEST-2, sped up recovery, restoring 80% of services by mid-morning.
Global cloud infrastructure scale
AWS serves over four million enterprise clients, including startups, banks, and governments. The outage exposed vulnerabilities in digital supply chains, impacting sectors from retail to education.
In Brazil, platforms like Mercado Livre saw 20% more connection errors. Internationally, UK government sites, like the Revenue Service, faced limited access. The five-hour incident caused estimated losses in the millions.
User guidance
To minimize issues during outages, restart routers and check Wi-Fi connections. Update Amazon apps and monitor status at health.aws.amazon.com.
- Avoid complex Alexa commands during outages.
- Use offline modes on Echo devices for basic functions.
- Check Downdetector for real-time complaint tracking.
Cloud provider dependency
AWS holds 30% of the global cloud market, per HG Insights. Centralization in datacenters like US-EAST-1 increases systemic risks. Annual AWS reports highlight investments in resilience, like datacenter duplication. The swift recovery minimized long-term damage, but the incident underscores the need for diversified providers for greater resilience.