Massive AWS outage hits major sites and apps

Amazon Web Services glitched out hard yesterday, leaving millions of users staring at loading screens and wondering if it was just them (it wasn’t).

The impact was widespread

AWS is the world’s leading cloud infrastructure provider (~30% market share), making yesterday’s outage the largest internet disruption in over a year.

Downdetector, an outage-tracking website, said it received ~11 million user reports of outages affecting ~2,500 companies.

• Facebook, Snapchat, Reddit, and Amazon.com were unavailable for many users, while a wide range of financial services—including Coinbase, Fidelity, Robinhood, and Venmo—also reported disruptions.
• The outage affected communication tools like Slack and Signal, some airlines including United and Delta, media/entertainment sites like Disney+ and The New York Times, AI tool Perplexity, and videogames including Fortnite and Roblox.

Amazon said major connectivity issues were fixed as of roughly 6 pm CT yesterday, or ~15 hours after the outage initially began.

Behind the crash: Amazon says the outage stemmed from a technical update to a widely used AWS database service called DynamoDB, which knocked the service offline in Amazon’s Northern Virginia data centers used by much of the US East Coast.

Once those data centers went down, it started a chain reaction where other AWS services also failed and stayed offline for hours, even after the initial problem was fixed.

Bottom line: “When a major cloud provider sneezes, the internet catches a cold,” said Mike Chapple, a Notre Dame IT professor and former NSA computer scientist.