Microsoft has outlined three problems behind the global Microsoft 365 outage on November 12–13: a dormant software bug, redirected Teams traffic, and reduced gateway capacity, according to BleepingComputer. Together, they triggered a cascading failure that left more than 80% of its networking gateway servers unresponsive.
The disruption ran from November 12 at 16:20 UTC until November 13 at 16:10 UTC, according to Microsoft’s preliminary incident report as covered by Ars Technica. That’s nearly a full day of trouble for people depending on Microsoft’s cloud infrastructure.
For users, the consequences went well beyond an unavailable app. Outlook, Exchange Online, Teams, and Copilot were all disrupted, with inaccessible email and significant delivery delays. Many people, companies, and services were negatively impacted if their work or collaboration depended through those services.
How Three Problems Became One Global Failure
The first piece was an update deployed to Microsoft 365 networking gateway servers starting in July. According to Ars Technica, the change introduced a defect that could exhaust CPU and memory resources under particular conditions, leaving affected machines unable to respond.
The testing gap is particularly striking. Wccftech reports that an internal Windows process encountered lock contention under certain workloads. Microsoft’s labs primarily used weaker hardware with fewer CPU cores, while the defect was more likely to surface on newer, higher-core-count production machines.
The second piece was a traffic shift. About 15 hours before the outage, Teams traffic moved from an alternative gateway onto the primary gateway shared with Exchange Online. That increased both the load and the mix of requests reaching the affected infrastructure.
The third piece left less room to absorb that change: nearly 14% of the gateway’s server capacity was already offline because of an unrelated issue. Microsoft said the remaining capacity could accommodate that reduction, but the combined conditions triggered the bug.
As machines stopped responding, traffic flowed to working servers, which then became overwhelmed too. Safeguards meant to contain failures within regions also failed to stop the disruption from spreading worldwide.
Why Recovery Took So Long
Restarting servers didn’t provide a lasting escape. Accumulated requests overwhelmed machines as they returned, while the age of the faulty change complicated a conventional rollback. Engineers ultimately deployed a fix restoring previously functional connection-handling behavior.
Microsoft has since expanded capacity, monitoring, and traffic-distribution safeguards. The outlet also says the defect was specific to Microsoft 365 gateway infrastructure and doesn’t affect broader Windows deployments.
The explanation remains preliminary. Ars Technica reports that Microsoft classified the incident as a technical issue unrelated to cyber activity or AI agents. The next checkpoint is Microsoft’s final incident report, which Wccftech says is planned by November 20, with additional corrective measures expected.
Noobenheim News for Nerds