Global Users Still Locked Out of M365 Search as Fix Stalls

Reading Time : 2 minutes

A sprawling Microsoft 365 disruption that began Monday morning has now entered its second day, with search functionality across core services still impaired even as email traffic slowly returns to normal. The incident, which Microsoft traces to a misconfigured authentication layer shared by multiple cloud products, has left organizations unable to locate messages, files, and chat histories at a time when digital workflows depend on instant retrieval.

 

 

Microsoft engineers spent much of Monday testing and rolling out fixes to the underlying authentication infrastructure, a process that required careful, incremental restarts to avoid triggering wider failures. By late evening, mail flow in Exchange Online had largely stabilized, though the company warned that some tenants would continue to experience delays as backlogged queues drained. The gradual recovery offered relief to IT teams, yet it also underscored how a single configuration error can cascade through an ecosystem serving hundreds of millions of users.

 

 

Search remains the stubborn bottleneck. As of early Tuesday, queries in Exchange, SharePoint, OneDrive, and Teams still return incomplete or failed results, hampering productivity for knowledge workers who rely on these tools to surface critical information. Microsoft has acknowledged progress in telemetry data but has not committed to a timeline for full restoration, leaving administrators in limbo and raising questions about redundancy and failover design in one of the world’s most widely used productivity suites.

 

 

The episode follows a string of Exchange Online incidents earlier in 2026, including a June mail-flow outage and an April access disruption, prompting scrutiny of Microsoft’s change-management and monitoring practices. For enterprises that have migrated deeply into the Microsoft cloud, the lingering search failure is more than an inconvenience; it is a reminder of the operational fragility that can accompany centralized dependency, and a signal that resilience strategies must account for platform-wide authentication faults as much as for regional outages.

 

Bénédicte Lin – Brussels, Paris, London, Beijing, Seoul, Bangkok, Tokyo, New York, Taipei, Hong Kong
Bénédicte Lin – Brussels, Paris, London, Beijing, Seoul, Bangkok, Tokyo, New York, Taipei, Hong Kong

 

#Microsoft365 #CloudOutage #ExchangeOnline #SearchFailure #TechDisruption #EnterpriseIT #SaaS #DigitalWorkplace #CyberResilience #CloudComputing