AWS Went Down — and Took the Internet With It

The internet was brought to its knees on Monday when Amazon Web Services experienced a massive, day-long outage. What started as a routine technical failure inside AWS’s northern Virginia data center quickly devolved into chaos, disrupting the lives of millions around the world. Banking customers were locked out of their accounts, travelers couldn’t check into their flights and smart home devices like Alexa suddenly went haywire. 

Amazon says its systems have largely been restored, but the damage was done. Some experts estimate the global financial impact could reach hundreds of billions of dollars, as businesses lost sales and workers sat idle waiting to get back online. It’s a stark reminder that the cloud isn’t an abstract idea — it’s the backbone of modern life, underpinning nearly everything we do.

This wasn’t a one-off event either. In fact, it’s the third major outage in five years tied to this AWS data cluster (known as US-EAST-1), which powers much of the internet’s traffic. And it comes just over a year after the infamous 2024 Crowdstrike malfunction, which hobbled hospitals, airports and banks around the world for days.

This latest incident is yet another example of just how fragile the world’s digital infrastructure really is. The systems that keep global commerce, communication and transportation running are concentrated in the hands of a few tech giants. And when one of them falters — even for a few hours — just about everything else does, too.

More Like ThisInside the Multi-Billion-Dollar Infrastructure Deals Powering America’s AI Boom

First, What Is AWS Exactly?

Amazon Web Services is Amazon’s cloud computing division. It functions as a kind of scaffolding for the internet, providing the storage, networking and computing power that keep millions of websites, apps and digital services up and running. 

Instead of building and maintaining their own on-site servers, companies rent what they need from AWS, scaling up or down as needed. It’s what enables a startup to launch a new app overnight, or a global enterprise to operate without owning a single physical data center. This model has become the foundation for how most of the digital world operates today, from streaming platforms to e-commerce sites to financial services.

Why Was AWS Down?

Amazon hasn’t provided full details about what triggered the latest AWS outage, but a picture of what happened is starting to take shape.

The disruption originated in the company’s US-EAST-1 region — its oldest and largest data center, located in northern Virginia. According to Amazon, the problem began with a malfunction in an internal subsystem that monitors the health of the network load balancers within its Elastic Compute Cloud (EC2) service. These load balancers are responsible for distributing traffic across servers to keep systems stable and responsive. When the monitoring system failed, it caused a chain reaction that affected how new network traffic was managed across the region.

As the EC2 internal network degraded, other AWS services started to break down with it. One of the most critical points of failure involved the domain name system (DNS), which translates website names into numerical IP addresses so computers can connect to the right servers. Among other things, the DNS issue affected Amazon’s DynamoDB API, a cloud database used to store user data and application information. When the DNS couldn’t correctly resolve the DynamoDB endpoint, all the applications that relied on it failed to respond, causing internet-wide failures.

All told, DynamoDB underpins more than 100 other AWS services, so the outage rippled through the entire ecosystem, slowing down or completely knocking out major services, including Connect, Lambda, Config and Amazon Bedrock. And so, businesses that depend on these tools were forced offline or left with limited functionality for much of the day.

By Monday evening, Amazon said it had mitigated the DNS issue and restored most of its core services. But the recovery process was gradual — several services continued to experience significant delays as queued requests were processed.

These kinds of errors happen for a number of reasons. Most of the time it’s the result of a routine maintenance issue or an unexpected server failure. Occasionally, a simple configuration mistake by an engineer can be the culprit. In rarer cases, an outage may be the result of a cyberattack, although there’s no indication that was the cause this time around.

Related ReadingAWS vs. Azure: A Guide

Who Was Affected?

Thousands of companies were affected by the AWS outage, ranging from finance apps like Venmo, Coinbase and Robinhood to streaming services like Hulu and HBO Max. Uber and Lyft users couldn’t book rides, AT&T and Verizon customers couldn’t send texts, and popular work tools like Slack and Zoom were malfunctioning.  

And of course, several of Amazon’s own services were hit, including its e-commerce site, Prime Video, Ring and Alexa. Within the company itself, warehouse and delivery workers reported on Reddit that several internal systems were down, disrupting operations.

The New York Metropolitan Transportation Authority (MTA) and the U.K’s HM Revenue & Customs website were among the public services that users could not access amid the outage as well.

On That Note . . .More Cloud Companies to Know

Why Is This Outage Such a Big Deal?

This certainly isn’t the first time big swaths of the internet have gone dark due to a cloud disruption — not even this year. Microsoft Azure and Google Cloud have experienced outages recently, too, and they likely will again.

What made this particular outage so significant is AWS’s size and scope. It is the largest cloud computing provider in the world, controlling about 30 percent of the market. Only two other companies come anywhere close — Microsoft and Google — and even they trail behind by a significant margin. In other words: Most of the internet runs on AWS. When it goes down, all the apps, websites and back-end systems that depend on it are at risk of going down, too.

Frequently Asked Questions