AWS DynamoDB DNS failure takes down much of the internet
A DNS automation race condition in us-east-1 cascaded globally for about 15 hours. Slack, Snapchat, Fortnite, Venmo and Xbox went down, with roughly 70,000 organisations affected.
- Service
- AWS us-east-1
- Date
- 20 October 2025
- Cause
- DNS
- Duration
- 14.5 hours
- Status
- resolved
- Severity
- critical
- Region
- Global
- Source
- Network World / ThousandEyes ↗
Cite this
Nuffield Consulting, “AWS DynamoDB DNS failure takes down much of the internet”, The Daily Tech Wire outage & breach log, 20 Oct 2025. https://techwire.nuffieldconsulting.com/incidents/2025-10-20-aws-us-east-1-outage/ (retrieved 15 Nov 2026).
This entry summarises public reporting. Please also cite the original source: Network World / ThousandEyes, https://networkworld.com/article/4124642/top-11-network-outages-and-application-failures-of-2025.html.
Related incidents
- 7 May 2026AWS us-east-1 zone down 28 hours after cooling failureAWS us-east-1 · outage · Power / cooling · 28 hours
- 29 Oct 2025Azure Front Door config change breaks Microsoft 365, Xbox and retailersAzure Front Door (global) · outage · Configuration change · 8.4 hours
- 9 Oct 2025Azure Front Door edge sites crash across EMEAAzure Front Door (EMEA) · outage · Software bug
- 2 Oct 2025Commonwealth Bank app, website and ATMs fail togetherCommonwealth Bank · outage · Software / dependency · 2 hours