A whole host of Microsoft 365 services failed late on Thursday – with Redmond blaming a “a portion of dependent service infrastructure in the North America region [that] isn't processing traffic as expected.”
The incident played out over 10 hours before Microsoft confirmed a fix.
Emails, Teams, Microsoft Defender XDR, and the Microsoft 365 admin center all suffered severe “service degradation” (largely stopped working) for thousands of users – as Microsoft worked to “refine our load balancing configurations to remedy residual imbalances across the environment.”
Even print jobs were affected for many users.
The incident comes six weeks after infrastructure handling traffic also backed up heavily for UK users, forcing Microsoft to reboot hardware – it blamed that extended December 9, 2025 incident on a “recent policy change that impacted service traffic balancing” without sharing details.
A combination of buggy updates and stretched infrastructure continue to plague Microsoft in terms of availability. Frustratingly for users, its status page also seemed to fail with the outage, suggesting that it was running on the same affected infrastructure (or potentially just overloaded due to users checking in.)
Infrastructure can, of course, fail for multiple reasons.
Microsoft’s string of recent Azure post-mortems gives some colourful examples. Among the most recent causes of incidents have been the “emergency ‘power off’ safety system in a portion of an impacted datacenter activating unexpectedly, “removing power to a subset of compute and storage infrastructure racks” within its West US 2 AZ.
Another recent incident affecting Azure Government customers was due to “an inadvertent automated key rotation [that] resulted in ARM failures to fetch authorization policies that are needed to evaluate access” [to ARM itself, or Azure Resource Manager – the gateway for management operations for Azure services.”
The link has been copied!
