By Interestana AI Editorial — AI-drafted, human-overseen. How we report
Microsoft 365 Outage Caused by Maintenance Bug
Microsoft announced on Thursday that a bug within its automated network maintenance request system was responsible for a massive outage that disrupted Microsoft 365 and Azure services. The system mistakenly removed Internet Protocol (IP) routes from an excessive number of devices, leading to the widespread service interruption. This incident affected numerous users and businesses relying on Microsoft's cloud infrastructure.
The company stated that the automated system was intended to perform routine maintenance tasks. However, a flaw in its execution logic caused it to over-apply route removals, effectively disconnecting a significant portion of the network. Engineers worked to identify and rectify the issue, restoring services as quickly as possible. The exact duration of the outage and the full extent of its impact are still being assessed by Microsoft and its affected customers.
This event highlights the complexities and potential vulnerabilities associated with large-scale automated infrastructure management. While automation offers efficiency, it also introduces risks if not rigorously tested and monitored. Microsoft has indicated that it is reviewing its maintenance protocols and system safeguards to prevent similar incidents from occurring in the future. The company has not yet provided a specific timeline for when all services were fully restored, but confirmed that remediation efforts were underway throughout Thursday.
Customers experienced disruptions across various Microsoft 365 applications, including Outlook, Teams, and SharePoint, as well as Azure cloud services. The scale of the outage underscores the critical dependence of many organizations on these platforms for their daily operations. Microsoft has committed to providing further details and post-incident analysis once its investigation is complete. The company is also expected to address customer concerns regarding service availability and reliability.
Original source — read the full reporting at the publisher:
Read on BleepingComputerGet the weekly AI digest
AI news + new model releases, weekly. Drafted by our agents, reviewed by humans.