On October 20, 2025, a sudden outage in Amazon Web Services’ (AWS) Northern Virginia region—known as US-EAST-1—sent shockwaves across the internet, disrupting everything from workplace chat apps to social media giants. The incident exposed just how deeply the world’s digital infrastructure relies on Amazon’s cloud, and reignited urgent questions about resilience in the age of hyperscale computing.
A Morning of Mayhem: What Happened?
At approximately 07:55 UTC, monitoring firm ThousandEyes detected widespread issues in AWS’s US-EAST-1 region, a critical hub for cloud operations in North America. The outage quickly cascaded, impacting
major services including Slack, Atlassian, Snapchat, and others. Users reported timeouts, error messages, and outright service failures as backend infrastructure buckled.
Amazon attributed the root cause to
DNS resolution issues affecting DynamoDB APIs—a core database service underpinning countless applications. Because so many AWS services and global features depend on endpoints in US-EAST-1, the disruption rippled far beyond the region itself.
Who Was Hit? The Domino Effect
The outage didn’t just inconvenience a few techies. It
knocked out access to workplace tools, social media platforms, and even some financial and airline systems. For businesses running critical workloads on AWS, the downtime meant lost productivity, frustrated customers, and in some cases, real financial losses.
Here’s a snapshot of the impact:
-
Slack: Messaging and collaboration ground to a halt for many teams.
-
Atlassian: Project management and developer tools went offline.
-
Snapchat: Users faced connectivity issues and failed message deliveries.
-
Other dependent services: Reports surfaced of issues with Netflix, Disney+, and more, though not all were directly confirmed in this incident.
How Long Did It Last?
The disruption lasted nearly two hours. ThousandEyes observed
service degradation until about 09:22 UTC, with full recovery by 09:35 UTC. AWS’s own Health Dashboard confirmed ongoing investigations and gradual restoration of affected services throughout the morning.
Why Does This Keep Happening?
This isn’t the first time AWS has suffered a high-profile outage, and it likely won’t be the last. Despite Amazon’s promises of “five nines” (
99.999%) uptime,
cloud outages remain a recurring risk—often due to the sheer complexity and interdependence of modern cloud architectures.
Experts and industry insiders have long warned about the dangers of putting all your digital eggs in one basket. As one cloud architect put it on Hacker News, “It’s a lot better to keep your architecture cloud-agnostic and test restores regularly on a different provider or region”. But for many companies, the cost and complexity of true multi-cloud resilience remain prohibitive.
Lessons Learned—and What’s Next
The October 20 outage is a stark reminder:
cloud convenience comes with hidden risks. For businesses, it’s a wake-up call to:
-
Review disaster recovery plans and ensure backups are stored outside primary cloud providers.
-
Consider multi-region or multi-cloud strategies for mission-critical workloads.
-
Monitor cloud service health dashboards and have contingency plans for communication and operations.
For AWS, the pressure is on to improve transparency and resilience, especially in its most critical regions. As the backbone of the modern internet, even a brief stumble can have global consequences.
The Bottom Line
As our digital lives become ever more entwined with the cloud, outages like this are no longer rare events—they’re existential threats to business continuity. The question isn’t if another AWS outage will happen, but when—and whether we’ll be ready next time.
Sources
1. AWS Outage Analysis: October 20, 2025 - ThousandEyes
2. The cost of egress traffic is a very good reason for many ...
3. Service health - Oct 20, 2025 | AWS Health Dashboard | Global
4. Cloud Egress Costs - Hacker News
5. [PDF] Competition and Risk
6. [PDF] Leveraging Client-side Storage Techniques for Enhanced Use of ...
7. [PDF] The Financial Services of Apple, Google and other Big Tech
