A digital blackout
On October 20, 2025, a major problem at Amazon Web Services (AWS) made many popular apps and websites stop working for almost 15 hours. People could not use Snapchat, Reddit, Netflix, Spotify, or even Alexa.
The issue began early in the morning in the United States and quickly spread around the world.
More than 17 million people and over 3,500 companies were affected, making it one of the biggest internet failures ever.
What is AWS?
Amazon Web Services, or AWS, is one of the most important parts of the internet.
Instead of every company buying and maintaining its own computers, they rent storage and [[computing power::The ability of a computer that shows how strong and fast it is at doing tasks and solving problems.]] from AWS. This is called cloud computing. It is like renting a car only when you need it instead of buying one.
Over 1.4 million companies, including Netflix, NASA, Disney, and McDonaldâs, use AWS to run their apps and websites.
Almost 70% of the worldâs [[internet traffic::The amount of data moving between computers, websites, and apps on the internet.]] passes through AWS centers in Virginia, USA, so even a small mistake there can affect people everywhere.
What went wrong?
A small computer error, called a race condition, made two systems change the same data at the same time.
This broke something called the Domain Name System (DNS). DNS is like the internetâs phonebook that helps computers find websites.
The mistake deleted key records for DynamoDB, a [[database::A place where information is stored on a computer so it can be easily found and used.]] used by thousands of apps. When DynamoDB failed, it caused a [[domino effect::A chain reaction where one event causes a series of similar events to happen, one after another.]], crashing over 100 AWS services and thousands of apps depending on them.
How did it affect everyone?
The outage hit nearly every part of daily life.
Students couldnât access online learning platforms like Canvas, leaving classes and tests on hold.
Banks and payment apps such as Coinbase and Venmo went offline, stopping people from making transactions.
Gamers were locked out of Fortnite and Roblox, while social media users couldnât open Snapchat or Reddit.
Even government websites in the U.K. went down, and smart devices like Alexa and Ring doorbells stopped responding.
Some airlines and major news websites also faced disruptions, proving just how deeply AWS supports the modern world.
How was it fixed?
AWS engineers manually repaired the broken DNS records and limited new system requests so their [[servers::Powerful computers that store, process, and deliver data or services to other computers or users over a network.]] would not overload.
Slowly, websites started working again, and by evening, most of the internet was back.
Later, AWS added extra safety checks and stronger backup systems to stop this from happening again.
Lessons for the future
The incident showed how [[fragile::Easy to break or damage; not strong.]] the internet really is. When one big provider fails, millions of people are affected.
Experts say companies should use more than one cloud service and keep backup plans ready.
It was a real reminder that even the biggest tech systems need strong safety nets.
Quick revision
On October 20, 2025, a major AWS outage stopped apps like Snapchat, Netflix, Spotify, and Alexa for about 15 hours, affecting millions worldwide.
AWS is a cloud service that powers much of the internet. About 70% of global internet passes through its centers in Virginia.
A technical error in AWSâs system broke the DNS and DynamoDB, crashing over 100 AWS services and thousands of apps depending on them.
