Amazon Web Services (AWS) reported new connectivity issues in its cloud computing unit on Monday, October 20, 2025, after mitigating a prior global outage. The initial failure, starting around 3:11 AM Eastern Time, affected thousands of websites and apps worldwide, including Snapchat, Reddit, and banking platforms. The issue originated in the US-EAST-1 region data center in Northern Virginia due to DNS resolution problems in the DynamoDB API endpoint.
Users reported over 6.5 million incidents globally via Downdetector, with peaks in the United States and United Kingdom. AWS identified the root cause as failures in internal network traffic monitoring systems. Technical teams implemented initial mitigations, restoring most operations by 6:35 AM Eastern Time.
The outage underscored the reliance on major cloud providers, impacting critical services like aviation apps and financial transactions.
- Snapchat: Users faced blocks in accessing the mobile app.
- Reddit: Failures in loading pages and forums.
- Ring: Home security devices stopped sending notifications.
Platforms affected by the initial outage
Companies reliant on AWS infrastructure experienced varied disruptions during the morning. Streaming, gaming, and financial platforms reported connection errors for about three hours.
Delta and United Airlines reported issues with apps and websites, affecting bookings and check-ins. Coinbase, a cryptocurrency exchange, confirmed temporary unavailability of trading.
Entertainment and gaming services disrupted
Roblox and Fortnite, popular among gamers, faced online server outages. Thousands of players reported inability to access accounts or start matches.
The impact extended to messaging apps like Signal, which saw encryption and call issues. Game developers reported delays in content updates.
Banks and finance under technical strain
Global financial institutions faced issues with digital transactions. Lloyds Banking Group in the UK requested patience from clients while restoring login services.
Robinhood, an investment platform, suspended trading for minutes. These events occurred during peak transaction hours.
Venmo and other mobile payment apps reported transfer failures. AWS limited new EC2 instance launches to stabilize the network.
Technical details of ongoing resolution
AWS is working on multiple fronts to address accumulated request backlogs. Status dashboard updates indicate significant recovery in services like SQS and Amazon Connect.
Engineers are investigating traffic balancing issues in the US-EAST-1 region. Clients were advised to clear data caches and restart connections.
Evolution of the issue throughout the day
The initial failure stemmed from a DNS resolution error, propagating to dependent databases. Technicians applied parallel fixes, reducing error rates to normal levels.
Subsequent spikes occurred around noon Eastern Time, possibly due to residual overload. The company monitors 24/7 to prevent recurrences.
Preventive measures under discussion
Cloud providers are revisiting redundancy protocols after similar events. Experts recommend infrastructure diversification to mitigate centralized risks.
AWS accounts for about one-third of the global cloud market, with outages impacting over a thousand platforms. Past incidents, like in 2023, lasted similar durations.
- Diversification: Companies shift to multiple providers.
- Monitoring: Early detection tools in data centers.
- Backup: Local systems for critical services.
