Microsoft said Thursday evening that some of the servers knocked out by a daylong Microsoft 365 outage were back online, but the company had not declared the incident over and gave no time for a full restoration.
The outage, tracked as incident EX1550543, began Thursday morning and left users unable to reach Exchange Online mailboxes, hitting Outlook email, Teams, SharePoint Online, Microsoft 365 Copilot and Defender XDR. Downdetector recorded more than 130,000 Outlook problem reports around 1:27 p.m. ET, along with more than 46,000 reports for Microsoft 365.
Microsoft said a large number of its servers had become unexpectedly unavailable after failing automated health checks, putting extra strain on the systems that remained. The company said the elevated load was not caused by extra traffic, and that nothing it had published pointed to a cyberattack.
By late afternoon, Microsoft said it had validated a different way of running server health checks on a subset of affected servers, allowing those systems to return to service, and it began rolling the fix out more widely. At about 6 p.m. ET, the company said a portion of the servers was back in service and "stable so far," and it promised further updates as the recovery continued.
A separate Teams incident in North America earlier Thursday was declared fully resolved around midday. The wider outage also disrupted connectors used by OpenAI's and Anthropic's AI services, Reuters reported.
Microsoft paused all deployments, moved traffic off affected servers and began restarting them as it worked to restore service. The company has not published a root cause for the failure.
