Microsoft 365 outage takes down Exchange Online, Teams and Intune as fixes fall short

Microsoft blames an overwhelmed core process that makes servers fail health checks, and restarted servers have stopped responding again once traffic returns.

ComputingPublished
What happened
A Microsoft cloud outage beginning around 18:00 UTC on 12 November knocked out Exchange Online, Teams, Copilot 365 and Defender for Office 365.
What’s new
Microsoft says its recovery efforts have not worked as hoped, and restored servers become unresponsive again once they receive traffic.
Why it matters
Many customers cannot send or receive email, and administrators also lost Intune and other management tools they would use to respond.

A large outage hit Microsoft's cloud services at around 18:00 UTC on 12 November, according to The Register. Exchange Online, Teams, Copilot 365 and Defender for Office 365 became unavailable. Admin tools also went down, including Intune and the Microsoft 365 Apps Admin Center, which left administrators without a way to investigate. The Register reports that email has stopped flowing for many Microsoft customers.

In posts on X, Microsoft said affected servers were failing automated health checks and dropping out of service. Its early investigation pointed to a specific component that processes server health. Microsoft said it had found no system changes that could have triggered the incident. The Register notes that the company's status page appears to contradict this, because it says engineers are considering rolling back a small number of recent configuration updates.

Microsoft says an alternative health check method allowed some servers to return to service. According to its status page, however, a core process on the affected machines is becoming overwhelmed, causing them to stop responding and reducing capacity. The status page says the steps taken so far "have not provided the anticipated relief."

In an update reported by The Register at 00:35 UTC on 13 November, Microsoft said some servers had come back online but could become unresponsive again once traffic reached them. Microsoft said it is now testing a recovery approach on an isolated portion of its infrastructure, before rolling it out more broadly. The company has not given a timeline for full restoration, according to The Register.