Microsoft explained the reasons for the recent massive Xbox outage and promised to make the system more resilient
An internal service failed.
Microsoft has revealed new details about the massive Xbox outage that occurred in July and lasted for about 15 hours. Xbox CTO Scott Van Vliet stated that the problem was caused by the failure of an internal service responsible for verifying licenses for purchased games and Game Pass content.
According to Van Vliet, the same service is used by several Xbox features on PC. The situation was further exacerbated by a bug in the Xbox client's software stack: when licensing-related services stopped responding, other license checks also began to fail.
This resulted in a cascading failure that affected several related services. In particular, users could not even launch games they already owned. Thus, the problem turned out to be significantly more serious than a simple disconnection of individual online functions.
After the incident, Microsoft transferred responsibility for the problematic licensing services to its own internal team. Engineers are also working to improve the resilience of the entire system to such failures. The discovered bug in the software stack has already been fixed — an update with the corresponding fix began rolling out yesterday.
Van Vliet emphasized that the company does not want to limit itself to fixing a specific bug. Xbox engineering and operations teams are working to improve the overall reliability of the infrastructure, and Microsoft promises to reveal additional details about its plans in the coming weeks and months. In future updates, fixes for such problems will be released along with new features and service quality improvements.
At the same time, the CTO separately acknowledged that the outage itself was a serious problem, but the situation became even worse because the failure of one service triggered a chain reaction in other components of the system.
I'm less concerned about one line with the cause of the failure and more about the real questions: why did the failure of one service manage to take down such a large part of the system, why did recovery take so long, and what will we change so that a single point of failure can no longer ruin your evening.
According to Van Vliet, Microsoft intends to strengthen dependencies related to authorization and game launching, improve the detection and localization of such failures, and inform users about emerging problems more quickly and clearly.
Комментарии