Microsoft confirmed on Thursday that its Azure services have been fully restored following a widespread global outage on Wednesday. The disruption affected not only Microsoft-owned platforms, including Microsoft 365, Outlook, Xbox Live, and Minecraft, but also external services such as Alaska Airlines and Starbucks. The company attributed the outage to issues with Azure Front Door (AFD), its cloud content delivery network (CDN) used by numerous services worldwide.
Shortly after the outage was reported, Microsoft released a statement on its X platform (previously known as Twitter), stating, "We are looking into a problem affecting multiple Azure services. Customers might encounter difficulties when trying to access these services."
Microsoft Azure services faced a disruption from 3:45pm UTC (9:15pm IST) on Wednesday, October 29 until 12:05am UTC (5:35am IST) on October 30. According to the company on a support page, customers and Microsoft services utilizing AFD encountered delays, timeouts, and errors.
The outage monitoring service Downdetector began observing issues with Azure services at 9:17 PM IST, and the number of complaints reached its highest point at 9:47 PM IST, with over 16,000 reports. Problems were noted with both server connections and domains.
According to Microsoft, the impacted services were App Service, Azure Communication Services, Azure Virtual Desktop, Microsoft Defender External Attack Surface Management, Microsoft Purview, and Microsoft Sentinel.
Microsoft attributed the outage to an accidental configuration change in Azure Front Door (AFD), its global content delivery network. The misconfiguration caused numerous AFD nodes to fail, resulting in widespread latency, timeouts, and connection errors that affected both Microsoft services and customer applications relying on AFD.
Microsoft swiftly halted any further configuration modifications and reinstated the most recent stable setup throughout its worldwide Azure Front Door network. The recovery process included reloading nodes and progressively redistributing traffic to return to normal operations. Customer modifications to AFD are temporarily on hold, and the company has introduced extra safeguards and validation measures to avert future problems. A comprehensive Post Incident Review will be provided to affected customers within 14 days.