Between 15:41 UTC on 29 October and 00:05 UTC on 30 October 2025, an issue within Azure Front Door (AFD) and Azure Content Delivery Network (CDN) caused global connectivity problems. This led to intermittent connection timeouts and DNS resolution errors, affecting access to some Cambrionix Connect services hosted via Azure Front Door.
During the disruption, some users experienced slower responses, connection errors, or temporary unavailability when accessing our cloud services. As Microsoft mitigated the problem, service availability gradually improved from around 18:30 UTC, returning to normal by 00:05 UTC.
Microsoft determined that a sequence of valid customer configuration changes on Azure Front Door generated incompatible metadata across two system versions. This exposed a latent defect in AFD’s data-plane software, causing crashes across global edge sites and resulting in the widespread timeouts and DNS failures.
Microsoft has since fixed the control-plane and data-plane defects, added new validation stages, and increased safety checks to prevent recurrence.
We closely monitored the situation in collaboration with Microsoft’s Azure engineering team.
Failover mechanisms were engaged where possible to minimize disruption.
Continuous service monitoring confirmed full recovery once Azure restored global stability.
We continue to track Microsoft’s post-incident improvements to Azure Front Door and will update customers if any additional mitigations are required on our side.
For Microsoft’s full RCA and mitigation roadmap, see: https://aka.ms/air/YKYN-BWZ
No components marked as affected
Resolved
Between 15:41 UTC on 29 October and 00:05 UTC on 30 October 2025, an issue within Azure Front Door (AFD) and Azure Content Delivery Network (CDN) caused global connectivity problems. This led to intermittent connection timeouts and DNS resolution errors, affecting access to some Cambrionix Connect services hosted via Azure Front Door.
During the disruption, some users experienced slower responses, connection errors, or temporary unavailability when accessing our cloud services. As Microsoft mitigated the problem, service availability gradually improved from around 18:30 UTC, returning to normal by 00:05 UTC.
Microsoft determined that a sequence of valid customer configuration changes on Azure Front Door generated incompatible metadata across two system versions. This exposed a latent defect in AFD’s data-plane software, causing crashes across global edge sites and resulting in the widespread timeouts and DNS failures.
Microsoft has since fixed the control-plane and data-plane defects, added new validation stages, and increased safety checks to prevent recurrence.
We closely monitored the situation in collaboration with Microsoft’s Azure engineering team.
Failover mechanisms were engaged where possible to minimize disruption.
Continuous service monitoring confirmed full recovery once Azure restored global stability.
We continue to track Microsoft’s post-incident improvements to Azure Front Door and will update customers if any additional mitigations are required on our side.
For Microsoft’s full RCA and mitigation roadmap, see: https://aka.ms/air/YKYN-BWZ