AWS Outage Explained: How a DNS Failure in Northern Virginia Knocked Major Apps Offline on 20 October 2025
A widespread disruption on 20 October 2025 left millions of people unable to access some of the internet’s most familiar services. The incident originated at Amazon Web Services (AWS)—a U.S. cloud giant that underpins a large share of global web traffic—and cascaded across social media platforms, banks, and popular games. Here’s what went wrong, who was hit, and why the impact felt so big.
What Happened — And Where the Failure Began
AWS confirmed that the outage was tied to a Domain Name System (DNS) issue in its Northern Virginia (US-EAST-1) data-centre region, one of its oldest and busiest sites. DNS acts like the internet’s address book: it translates names into the machine-readable locations that apps use to reach servers. When that “map” fails, apps and websites are still running—but devices don’t know where to find them.
Plain-English takeaway: users tapped apps and clicked links, but the traffic couldn’t be routed correctly, so services appeared “down.”
Who Was Affected — A Snapshot Across Sectors
The outage touched a broad set of brands and industries, highlighting how many organisations rely on AWS as core infrastructure.
| Sector | Examples Reported |
|---|---|
| Social & Community | Snapchat, Reddit |
| Banking & Finance | Lloyds, Halifax |
| Gaming | Roblox, Fortnite |
| Creative & Productivity | Canva |
| Government & Public Services | HMRC |
Impact scale: 1,000+ companies experienced disruption, affecting millions of users.
Why a DNS Glitch Breaks So Much, So Fast
When DNS “loses its bearings,” the problem spreads quickly because downstream systems depend on correct lookups to reach databases, storage, and APIs. Even if the underlying apps are healthy, failed DNS resolution means login requests, content feeds, or payment calls can’t reach the right endpoints—so pages time out and apps won’t load.
Common triggers for such errors include:
-
Routine maintenance gone awry
-
Server or component failure
-
Human error (misconfiguration)
-
Cyber attack (no evidence of this in the current incident)
Why the Internet Feels Centralised — And the Limited Alternatives
AWS has become a backbone provider by offering compute power, storage, databases, and traffic management at global scale. Critics warn that concentrating so much infrastructure with one vendor creates a “single basket” risk. The challenge: few rivals can match AWS’s breadth.
-
Main alternatives at similar scale: Microsoft Azure, Google Cloud Platform
-
Smaller competitors: IBM, Alibaba
-
Newer European option: Stackit (launched by the parent company of Lidl)
Despite fresh competition, AWS remains the dominant player, which is why a problem in one region can feel like the entire internet is wobbling.
Key Facts at a Glance
| Item | Detail |
|---|---|
| Incident date | 20 October 2025 |
| Primary region | Northern Virginia (US-EAST-1) |
| Technical cause | DNS error disrupting routing to services |
| Scale | 1,000+ companies, millions of users affected |
| Notable services | Snapchat, Reddit, Roblox, Fortnite, Lloyds, Halifax, Canva, HMRC |
| Security note | No evidence of a cyber attack reported |
The Bigger Question Raised by the Outage
Experts point to a simple lesson: the more businesses rely on a single cloud provider, the greater the ripple when that provider stumbles. There are only a handful of platforms capable of running the modern internet at AWS’s scale, which means resilience conversations—across governments, banks, and consumer apps—are likely to intensify after this incident.
