Microsoft says a software flaw that sat unnoticed in its systems since July caused the worldwide Microsoft 365 outage that ran for nearly a full day last week, leaving more than 80% of its networking gateway servers unresponsive at the worst point.
The disruption began at 11:20 AM ET on Thursday, November 12, and affected Outlook, Exchange Online, Teams, and Copilot, according to the company. Users around the world lost access to email, and message delivery fell significantly behind. Microsoft said service was fully restored at 11:10 AM ET on Friday, November 13.
In a preliminary post-incident report published in the Microsoft 365 Admin Center, Microsoft attributed the outage to three factors: the latent bug, a traffic increase that followed a Microsoft Teams failover, and reduced server capacity. The bug entered Microsoft 365 networking gateway servers through an update built in July. That update ran for more than three months without trouble, but it could overwhelm servers under particular traffic conditions and cause them to stop responding. Microsoft said its lab tests missed the flaw because they relied on weaker hardware with fewer cores than its production machines, while the contention problem is worse on newer machines with more cores.
The bug surfaced after Microsoft Teams automatically moved onto the networking infrastructure used by Exchange Online roughly 15 hours before the outage, raising traffic on those servers. Teams also suffered a separate outage earlier that day, though Microsoft has not established whether the two events are connected. About 14% of gateway server capacity was already offline because of an unrelated issue. Microsoft said the remaining capacity was sufficient on its own, but the added traffic and the lost servers together set off the defect.
As machines failed, Microsoft's systems sent users to other servers, which were overwhelmed in turn and also went down. Microsoft said the unknown defect spread across its fleet in under two hours and took more than 80% of servers offline at the low point. Safeguards meant to keep failures from crossing regional boundaries did not hold, and the outage spread worldwide. Teams, OneDrive, and SharePoint moved to alternate infrastructure and recovered sooner, but Exchange Online had no such option, so mailboxes stayed unreachable. Restarts were difficult because returning servers were immediately swamped by queued traffic.
Engineers wrote a fix once they identified the bug and pushed it to affected servers, and services stabilized over Friday. Microsoft said it has since added capacity, tightened monitoring, and put new safeguards in place against traffic redirection that could cause similar cascading failures. Charles Lamanna, Microsoft's executive vice president of Copilot, Agents, and Platform, said in a statement on X that the company is deeply sorry for the effect on customers and their business operations. Microsoft said its investigation continues and a final report is expected later this week.
More hardware news from TechManNews.






