Amazon Web Services suffered a major incident centered on its US-EAST-1 region in Northern Virginia early Monday, knocking popular apps and websites offline and driving error rates and latency higher across multiple services.
AWS reported recovery underway by 5:27 a.m. ET and said most requests were working again by about 6:35 a.m. ET.
The timeline
AWS first acknowledged problems in US-EAST-1 at 3:11 a.m. ET. In subsequent posts it said engineers were “actively engaged” on mitigation, then noted “significant signs of recovery” at 5:27 a.m. ET and “most AWS Service operations are succeeding normally” at 6:35 a.m. ET. Those updates align with coverage that tracked status-page messages through the morning.
What went down
The disruption rippled across social, gaming, finance, communications, and Amazon’s own products. Services affected included Snapchat, Fortnite, Roblox, Signal, Coinbase, Robinhood, Venmo, Prime Video, and Alexa, among others, according to outage trackers, company posts, and media tallies.
Several collaboration and AI tools also saw issues, including platforms that run on AWS cloud infrastructure. Coverage noted problems for Perplexity, Airtable, Canva and others that rely on Amazon’s data centers.
U.S. and U.K. knock-on effects
The impact was global. In the U.K., banks such as Lloyds, Halifax, and Bank of Scotland reported issues, and government services including HM Revenue & Customs experienced disruption, per incident trackers and local reporting. Telecom providers Vodafone and BT were also cited among those with problems.
Early coverage also flagged airline websites in the United States, including Delta and United, among services showing problems during the window.
What AWS said
Amazon’s status messages pointed to increased error rates and latencies in US-EAST-1 and later indicated broad restoration as mitigations took effect. The company emphasized recovery while it continued to work through backlogs of queued requests.
Was DynamoDB the culprit?
DynamoDB, AWS’s managed NoSQL database, showed degraded performance during the incident. Some outlets linked the impact to DynamoDB, with one summary quoting an AWS status note about DNS resolution issues affecting the DynamoDB API endpoint in US-EAST-1.
AWS has not published a formal post-event analysis confirming a single root cause, so DynamoDB should be treated as implicated but not definitive at this stage.
Context and engineering takeaways
US-EAST-1 is among AWS’s most heavily used regions, hosting critical workloads for consumer apps, enterprises, and public-sector systems. Outages there tend to produce outsized effects well beyond North America.
Monday’s event echoes high-profile disruptions in 2020–2023 and will likely renew customer focus on multi-Region architecture, automated failover, and resilience testing.
For regulated industries and public services, incident reviews typically revisit single-Region dependencies, cross-cloud strategies, and DNS hardening to reduce blast radius in future faults.
