Microsoft details how three problems combined for a global outage
A hidden bug, shifting traffic, and reduced capacity led to cascading failures.
https://arstechnica.com/gadgets/2026/11/microsoft-details-how-three-problems-combined-for-a-global-outage/?utm_brand=arstechnica&utm_social-type=owned&utm_source=mastodon&utm_medium=social
@arstechnica
Do I understand correctly that the underlying bug deployed in July could have been discovered if testing had been performed on a hardware clone of the production environment?