A flood of API calls has caused overload of route reflectors which led to the degradation of connectivity. The source has been identified and isolated, followed by a restart of the subsystem in question. Connectivity is now fully restored in all services.
Posted May 19, 2025 - 02:15 UTC
Update
The impact is limited to newly created load-balancers in Managed Kubernetes clusters and instances of Managed PostgreSQL and MLflow. Existing instances of LBs and managed services might experience problems in updating target statuses.
Compute VMs and other services should not be affected.
Posted May 19, 2025 - 01:17 UTC
Investigating
Some customers might experience connectivity issues when working with resources in eu-north1 region. We are investigating the issue and clarifying the impact.
Posted May 19, 2025 - 00:20 UTC
This incident affected: EU-NORTH1 (Virtual Private Cloud (Networking)).