Amazon says AWS S3 servers in Northern Virginia are now operating normally after earlier outage, which impaired sites including Quora, Trello, IFTTT, and Imgur
An outage of Amazon Web Services' Simple Storage Service (S3) impacted sites across the internet Tuesday.
Context & Ripple Effects
Tuesday's S3 failure is the template for a pattern that keeps repeating: when the Simple Storage Service in Northern Virginia goes down, large swaths of the consumer web go with it, because Quora, Trello, IFTTT, Imgur and countless others build on the same regional endpoint. Amazon's 'operating normally' status closes the incident, but it is also the first entry in what related coverage shows is a series — another North America-wide AWS outage hit Roku, Flickr and Adobe in November 2020, and a December 2021 disruption took down Disney+, Tinder, Coinbase and Venmo.
First-order effects
- Quora, Trello, IFTTT, Imgur and other S3-dependent services regain full functionality as Amazon restores the Northern Virginia servers, ending hours of broken images, feeds and integrations across the web.
Second-order effects
- Each recurrence pushes AWS customers toward multi-region and multi-provider architectures, and hands Microsoft and Google a standing argument for workloads that treat any single region — including US-EAST-1 — as a failure domain rather than a home.
Third-order effects
- The 2017 incident foreshadows the structural problem confirmed by the October 2025 US-EAST-1 outage: enormous portions of the internet concentrate on one AWS region, making its availability a systemic risk that regulators and enterprise buyers increasingly have to price in.
The trend: Cloud computing's efficiency concentrates critical infrastructure in a handful of regions, so a single provider's regional failure now produces internet-scale outages on a repeating cycle.