Managed Kubernetes: Service Degradation for Control Planes in FRA

Update timeline

1. investigating Mar 31, 2026, 12:22 PM UTC

   We are currently investigating service degradation for Control Planes of some Kubernetes clusters in the FRA DC.

2. monitoring Mar 31, 2026, 01:10 PM UTC

   All affected Control Planes have recovered. We are currently monitoring the affected clusters for further anomalies.

3. resolved Apr 01, 2026, 08:18 PM UTC

   We are marking the incident as resolved, as no further anomalies were detected. The root cause was a transient latency issue in one of the redundant CoreDNS pods, which led to kube-apiservers being unable to discover etcd instances in time. We are currently developing better mitigation options for the excessive NXDOMAIN requests observed during the incident.