Amazon Technologies – a Major AWS Outage in October 2025

In October 2025, a significant AWS outage occurred due to internal issues affecting DNS resolution and load balancer health checks. The incident originated in the US-EAST-1 region and resulted in global disruptions across various services.

While the outage was resolved within a few hours, it highlighted the fragility of cloud-reliant systems and the importance of architectural redundancy.

Page Index

• Key Aspects
• Regional Service Dependence
• DNS Resolution Failure
• Load Balancer Health Checks
• Widespread Service Disruption
• Lessons in Cloud Resilience
• Conclusion
• Why half the internet just went dark – 3 mins
• US-EAST-1 is humanity’s weakest link – 5 mins

Key Aspects

• The outage originated from the US-EAST-1 region, a critical AWS data center location.
• DNS resolution problems disrupted communication with core services, such as DynamoDB.
• Health monitoring systems for load balancers experienced internal faults.
• Many globally used apps and websites were affected due to regional reliance.
• The incident highlighted the importance of resilient cloud architecture and DNS strategies.

Regional Service Dependence

The AWS outage was centered in the US-EAST-1 region, one of the most widely used AWS regions for hosting and deploying cloud infrastructure. Many organizations select this region for its low latency and broad availability of services, making it a common default choice. However, this widespread use created a single point of failure when issues arose in the region.

Because a large number of critical workloads and services operate out of US-EAST-1, disruptions there have a disproportionately large impact. The incident highlighted the potential consequences of relying heavily on a single AWS region, which can lead to global service disruptions affecting everything from retail websites to social media and financial platforms.

DNS Resolution Failure

A key component of the outage was a failure in DNS (Domain Name System) resolution, which prevented applications from translating service names, such as those used by Amazon DynamoDB, into usable IP addresses. When DNS fails, even healthy services cannot be reached, creating widespread breakdowns across dependent applications.

The issue was traced to internal DNS resolution problems within AWS’s EC2 network. As DNS is foundational for service discovery and communication, this failure had a cascading effect across various systems, demonstrating how tightly interwoven cloud service dependencies can be.

Load Balancer Health Checks

Another contributor to the outage was a malfunction in the subsystem responsible for monitoring the health of AWS’s network load balancers. These health checks are crucial for directing traffic only to functional resources. When the monitoring system failed, it triggered incorrect traffic routing and amplified service instability.

The error led to false assumptions about resource availability, compounding access issues to services like DynamoDB. AWS’s internal monitoring and routing systems are designed for high resilience, but this incident showed that failures in automated checks can have outsized effects across a distributed system.

Widespread Service Disruption

The outage disrupted major applications, including Reddit, Snapchat, Venmo, and banking apps, all of which rely on AWS infrastructure. Because these services rely on high availability from cloud providers, any incident affecting one region can have a global impact, particularly if failover configurations are inadequate or underperforming.

Many organizations learned that depending heavily on a single cloud region or service introduces systemic risk. The outage underscored the importance of multi-region architectures and robust failover planning, even for companies that otherwise operate primarily within a single AWS availability zone.

Lessons in Cloud Resilience

The October 2025 AWS outage highlighted critical lessons in cloud architecture, especially around resilience planning. It demonstrated the necessity of distributed deployment, DNS failover strategies, and proactive monitoring of internal service dependencies. For IT organizations, it reinforced the importance of incident response protocols that extend beyond immediate fixes to include service backlog management and full system recovery.

Organizations must prepare for indirect failures—where a non-obvious component, such as internal DNS, can cripple broader operations. Investing in cross-region redundancy and scenario-based disaster recovery planning can help mitigate similar risks in the future.

Conclusion

The October 2025 AWS outage revealed how even well-architected cloud systems can experience critical failures. It clarified the need for IT teams to prioritize regional redundancy, DNS resilience, and comprehensive incident recovery strategies.

Why half the internet just went dark – 3 mins

US-EAST-1 is humanity’s weakest link – 5 mins