AWS is down. Amazon Web Services crashed this morning, taking down dozens of apps, games, and digital platforms. The outage began at 9 a.m. Italian time from data centers in Northern Virginia, the region us-east-1 which represents the beating heart of global cloud computing. Amazon has confirmed problems at DynamoDB e EC2, the two fundamental services that provide databases and computing power to thousands of companies. In just a few minutes Downdetector It has seen simultaneous spikes in reports for Snapchat, Roblox, Fortnite, Duolingo, Ring, Coinbase, Robinhood, and many others. This isn't an isolated case: when AWS infrastructure goes down, the ripple effect is immediate and global.
AWS Down: What Happened at the Northern Virginia Data Centers
Amazon has the problem was officially recognized on its status page, confirming “increased error rates” and delays that have affected several services in the region US-EAST-1. The problem originates in DynamoDB, AWS’s distributed NoSQL database, and EC2 (Elastic Compute Cloud), the compute service that powers millions of applications. These two pillars of cloud infrastructure provide data storage and computing power, respectively, to thousands of companies that rent AWS resources to run their applications.
The fact that completely different services went down simultaneously indicates a deep infrastructure problem, not an isolated failure. Northern Virginia isn't just any region: it's AWS's largest and busiest data center, the one that handles a significant portion of global internet traffic. When us-east-1 he has problems, the impact is felt everywhere.
The list of services ko
The list of affected services is long and varied. Among the most notable: Snapchat for social media, Roblox, Fortnite, Epic Games Store e Playstation Network for gaming, Duolingo e Canvas for education. Services within the Amazon ecosystem have also gone offline: RING (the security cameras), Amazon Alexa e Amazon Prime Video. On the financial front, problems for Venmo, Robinhood, Chemistry e Coinbase. Even Crunchyroll, the anime streaming platform, has stopped working.
Secondo a study by the Uptime Institute Published in April 2025, 31 such cloud incidents occurred in 2024. The research highlights how Applications built on multi-zone architectures experienced only 19 minutes less downtime than single-zone ones, demonstrating that geographic redundancy is often not enough when the problem is systemic.
AWS Down, the fragility of the giant
Northern Virginia has a recurring problem. According to StatusGator data, in the 2022 us-east-1 It recorded 23 partial outages for a total of 61 hours, representing 51% of all AWS outages and 68% of the overall outage duration. This is no coincidence: this region offers 215 different services and is the most used in the world, which means more traffic, more complexity, and more outage points.
The paradox is clear. We build "distributed" and "resilient" cloud architectures, but then we put everything in the same data center because it's convenient, fast, and cheap. Until it's not anymore. PagerDuty survey A survey of 1.000 IT and business executives revealed that 88% expect another major global outage in the next 12 months. That's not pessimism: it's statistics.
The real problem isn't that AWS is going down. The real problem is that when AWS goes down, everything else goes down too. As already discussed here on Futuro ProssimoThe concentration of digital infrastructure in the hands of a few giants creates systemic vulnerabilities that no amount of redundancy can truly address.
Amazon is working to restore services, but hasn't provided an estimate for a full resolution. Meanwhile, millions of users are waiting for Northern Virginia to get back online. And they're wondering, perhaps for the first time, just how fragile the invisible infrastructure that supports their digital lives really is.
