Overview
- Microsoft said an inadvertent configuration change in Azure Front Door triggered DNS and routing problems starting around 16:00 UTC, disrupting access to the Azure portal and services including Microsoft 365, Teams, Xbox, Minecraft and Copilot.
- Engineers blocked further Azure Front Door changes, failed the portal away from the affected infrastructure, rerouted traffic, recovered nodes and began rolling back to a stable configuration.
- User reports on Downdetector peaked at over 18,000 for Azure and nearly 11,700 for Microsoft 365 before trending downward as mitigation progressed, though full recovery was expected to take additional hours.
- Operational knock‑ons were reported by businesses and public services such as Alaska Airlines, Vodafone in the UK and Heathrow Airport, with Scotland’s Parliament suspending voting due to technical issues.
- The incident came days after a major AWS disruption and renewed concerns about reliance on a few hyperscale cloud providers, with experts warning of concentration risk and cascading failures.