Active - Multiple services experiencing connectivity issues in multiple regions

Starting at 20:30 UTC on 30 September 2026, customers using ExpressRoute and VPN Gateway may have experienced degraded or interrupted network connectivity. Customers may also experience failures or delays with some network management operations.Current status: ExpressRoute gateways are showing significant recovery as of 22:35 UTC, and we continue to monitor service health to validate that recovery is sustained. We have paused infrastructure servicing activity associated with the onset of this event, preventing additional resources from entering the affected workflow.For VPN Gateway, some affected gateways experienced reduced redundancy or loss of connectivity, and recovery continues to be monitored. Some supporting network management components have not recovered automatically and continue to contribute to management operation failures in affected regions. Engineering teams are actively restoring these components using healthy instances where available.Our investigation has identified a correlation between the timing of this event and infrastructure operating system servicing activity. The underlying failure mechanism has not yet been confirmed, and we continue to investigate why some network service instances became unhealthy and why recovery behavior differed across services.We will provide additional information within the next 30 minutes, or sooner if events warrant.

Starting at 20:30 UTC on 30 September 2026, customers using ExpressRoute and VPN Gateway may have experienced degraded or interrupted network connectivity. Customers may also experience failures or delays with network management operations across impacted services. Azure Firewall data-plane traffic is not currently known to be impacted. However, customers in affected regions may experience failures or delays when performing service management operations, including creating or updating Azure Firewall resources and updating firewall rules.As of 22:35 UTC on 30 September 2026, ExpressRoute gateways are showing significant recovery, and we continue to monitor service health to validate that recovery is sustained. We have paused infrastructure servicing activity associated with the onset of this event, preventing additional resources from entering the affected workflow.Some supporting network management components have not recovered automatically and continue to affect management operations and the recovery of dependent resources in some regions.Engineering teams have paused the infrastructure servicing activity associated with the onset of this event. We are restoring the affected network management components, reducing load to stabilize the service, and addressing constraints affecting our ability to add resources. We continue to validate recovery region by region and service by service.Our investigation has identified a correlation between the timing of this event and infrastructure operating system servicing activity. The underlying failure mechanism has not yet been confirmed, and we continue to investigate why network service instances became unhealthy and why recovery behavior differed across services.We will provide additional information within the next 30 minutes, or sooner if events warrant.

Starting at 20:30 UTC on 30 September 2026, customers using gateway services may be experiencing degraded or interrupted network connectivity. Customers may see gateways failing to load in the Azure Portal, and may also be experiencing failures or delays with network management operations across the following impacted services:ExpressRoute GatewaysAzure FirewallApplication Gateways / Web Application FirewallVPN GatewaysAzure VMWare Solution (AVS)Current status: The vast majority of previously-impacted regions are showing signs of recovery, we are working to apply configuration changes that were successful in other regions to the remaining five regions: France Central, North Europe, Southeast Asia, UK South, and UK West.Our investigation has identified a correlation between the timing of this incident and infrastructure operating system servicing activity, which has since been paused. The underlying failure mechanism has not yet been confirmed, and we continue to investigate why network service instances became unhealthy and why recovery behavior differed across services.We continue to validate recovery - region by region, and service by service. We will provide additional information within the next 60 minutes, or sooner if events warrant.