San Francisco: Amazon’s cloud computing service, Amazon Web Services (AWS), has returned to normal operations following a major outage that disrupted thousands of websites and applications worldwide, from Snapchat and Reddit to Venmo and Zoom.
The disruption, traced to AWS’s northern Virginia data center cluster known as US-EAST-1, was one of the most extensive internet interruptions since last year’s CrowdStrike malfunction that crippled hospitals, banks, and airports across several countries.
Amazon said the issue stemmed from a problem within its EC2 internal network, specifically a subsystem that monitors the health of network load balancers. The malfunction affected the Domain Name System (DNS), which helps applications locate servers and services, preventing them from accessing AWS’s DynamoDB cloud database used to store user information and critical data.
By 3 p.m. Pacific Time (2200 GMT), the company confirmed that all services had been restored, though some such as AWS Config, Redshift, and Connect still had message backlogs expected to clear within hours.
Update on AWS service event. For the latest information: https://t.co/CFaXbXxkaF pic.twitter.com/rlVha1pYJr
— Amazon News (@amazonnews) October 20, 2025
The outage knocked out digital tools for companies and individuals across continents, disrupting business operations, online payments, and communications. In Britain, Lloyds Bank, Bank of Scotland, Vodafone, BT, and even the HMRC website experienced disruptions, according to Downdetector. Across the world, millions of users struggled to access platforms like Venmo, Zoom, Reddit, Snapchat, and Roblox.
Even Amazon’s own platforms, Prime Video, Alexa, and its shopping site were affected. Gaming platforms including Fortnite, Clash Royale, and Clash of Clans also went offline, along with services such as Coinbase, Robinhood, and Lyft.
The northern Virginia site, AWS’s oldest and largest region, has been at the center of at least three major outages in the past five years, with previous incidents recorded in 2020 and 2021. Amazon did not provide details about why this particular cluster remains vulnerable to repeated failures.
According to AWS documentation, US-EAST-1 is the default region for many of its services, amplifying the impact whenever it goes down. Cybersecurity experts said the outage underscores the risks of global dependence on a handful of cloud providers. According to Ookla, operator of Downdetector, more than 4 million users reported problems during the incident.
