Microsoft 365 large-scale outage: Unable to access emails via Outlook, affecting multiple services such as Teams - ITmedia NEWS

On November 12 (U.S. time), a disruption occurred that prevented users from connecting to the email mailbox in Microsoft's cloud service "Microsoft 365" email service "Exchange Online." This affected not only the web version of "Outlook," but also the desktop application, mobile application, MAPI, ActiveSync, and REST connections. Microsoft has restored some of the servers and is applying similar measures to other servers, but has not announced the time for full recovery. Microsoft announced on January 13 at 2:54 AM Japan time via its official X account that it is investigating reports from users in the U.S. that they are unable to access Exchange Online mailboxes through multiple connection methods. Later, it revised its explanation to state that the impact could potentially affect users in all regions. On its status page, Microsoft listed Exchange Online and other services such as "Microsoft Teams," "SharePoint Online," "Microsoft 365 Copilot," "Microsoft Defender XDR," "Microsoft Planner," "Microsoft To Do," and "Microsoft Purview" as possibly affected. In particular, several functions such as data loss prevention (DLP), automatic labeling, and data classification in Purview are reportedly impacted. Initially, Microsoft explained that the issue was due to high load on parts of the infrastructure in multiple regions, which was thought to be causing connection errors. In a later X post, it clarified that the load was not due to increased traffic, but rather because a number of servers unexpectedly stopped responding, causing the remaining infrastructure to become overloaded. The servers had become unavailable due to failing automatic health checks. Microsoft initially stated that no changes had been made around the time of the outage, but later explained that during a re-evaluation, a few configuration changes were identified and that these could have contributed to the outage, and that they are investigating this possibility. Microsoft has previously taken measures such as temporarily halting all deployments, distributing connection traffic across the global infrastructure to disperse the load, and restarting the affected servers. In a post at 6:00 AM Japan time on January 13, it explained that it had verified an alternative method of health checking and found that some of the affected servers had been restored. Currently, it is applying this change more broadly and is preparing to implement an automatic recovery mechanism that continuously detects and restarts servers that have stopped responding. It stated that the next update will be posted by 7:30 AM Japan time on January 13. A situation has also arisen where it is difficult to access the "Microsoft 365 Admin Center" for administrators. Microsoft is urging administrators who cannot use the Admin Center to check the information on the public status page.