started · updated
Microsoft explains massive global outage caused by software error
Microsoft has released a preliminary technical report regarding a massive global outage that paralyzed services, primarily Exchange Online, for nearly 24 hours. The disruption began on Thursday, November 12, at 17:20 CET and was resolved by Friday, November 13, at 17:10 CET.
The company clarified that the collapse was not caused by a cyberattack, but rather by a combination of three factors involving networking gateways. A software update deployed in July contained a hidden error known as lock contention. This flaw only manifested under heavy loads on modern servers with high CPU core counts, causing the error to evade testing on weaker hardware.
The outage was triggered when a routine failover mechanism shifted Microsoft Teams traffic to the primary gateway used by Exchange Online, significantly increasing the load. This occurred while approximately 14% of gateway capacity was already offline due to an unrelated incident from the previous day, creating a domino effect that impacted Outlook, Teams, Copilot, SharePoint, OneDrive, and the Defender security platform.