What a way to start the week! A widespread Amazon Web Services (AWS) outage has knocked a massive chunk of the internet offline, affecting popular online services that rely on the company"s cloud infrastructure. Reports of the outage began to surface on Monday, October 20, 2025, around 8 AM UK time (midnight PT).
In AWS"s status page, the company attributed the issue to a failure in its US-EAST-1 region, located in North Virginia. Amazon identified a potential root cause for the error rates, stating the problem appears related to DNS resolution of the DynamoDB API endpoint.
This means that applications were struggling to find the correct internet address for one of Amazon"s core database services, which then caused a massive cascading failure across at least 67 other services. Amazon confirmed it was working on multiple paths to fix the problem and that it was seeing "significant signs of recovery", but warned that a backlog of requests could cause lingering delays.
If you do not know, DynamoDB is Amazon"s fully managed NoSQL service. Think of it as a gigantic, incredibly fast, and flexible database that thousands of companies build their applications on top of. It is designed to handle immense traffic without breaking a sweat, which is why a failure in this specific service has such a widespread and damaging effect. When it goes down, it takes everyone who depends on it down, too.
Here"s a list of popular websites affected so far, and it is a long one:
- Amazon"s own shopping site
- Fortnite
- Snapchat
- Disney+
- Prime Video
- Roblox
- Robinhood
- Coinbase
- Epic Games Store
- Canva
- Ring
- Duolingo
- The New York Times
- Lyft
- McDonald"s App
- United Airlines
- Signal
- Hulu
AWS is the largest public cloud provider, holding about a third of the market. Its infrastructure is relied on by everyone from small startups to massive enterprises like Netflix and Disney.
Obviously, this is not the first time the service has gone down and taken the rest of the internet with it. You might remember the great S3 outage of 2017, where an engineer"s typo took out the S3 storage service and broke websites for hours.
Other similar events include the Kinesis outage in 2020 and a pair of major disruptions in December 2021, all of which stemmed from the same US-EAST-1 region.
Thanks for the tip, Barney T!!!