RC RANDOM CHAOS

Three Cascading Failures Knocked Microsoft 365 Offline

· via Ars Technica

Original source

Three small failures combined to take down Microsoft 365

Ars Technica →

The outage originated from a defect in Microsoft’s cloud gateway that lay dormant since July. The bug escaped detection during testing because quality assurance labs ran less powerful hardware than some production systems; newer machines with higher CPU core counts provided conditions the bug needed to trigger.

Scheduled maintenance accelerated the failure. Microsoft shifted Teams traffic from a redundant gateway to its primary system as part of routine failover—a standard safety measure—about 15 hours before the outage. The primary gateway suddenly carried significantly more traffic in an unfamiliar configuration, overwhelming its resources.

As CPU and memory exhaustion set in, the gateway rejected new connections, causing cascading failures across Exchange and Microsoft 365. The 24-hour outage beginning November 12 affected organizations worldwide. Microsoft’s incident report confirmed the cause was technical infrastructure failure, not cyberattacks or artificial intelligence systems.

Read the full article

Continue reading at Ars Technica →

This is an AI-generated summary. Read the original for the full story.