kubernetes resource_error ai_generated true

Error: failed to get cpu utilization: unable to get metrics for resource cpu: unable to fetch metrics from resource metrics API

ID: kubernetes/hpa-missing-metrics-server

Also available as: JSON · Markdown · 中文
85%Fix Rate
90%Confidence
1Evidence
2024-01-20First Seen

Version Compatibility

VersionStatusIntroducedDeprecatedNotes
metrics-server v0.6.4 active — — —
Kubernetes v1.28.3 active — — —
Kubernetes v1.29.0 active — — —

Root Cause

Horizontal Pod Autoscaler (HPA) cannot query CPU/metrics because the metrics-server is not deployed or is not functioning correctly.

generic

中文

水平Pod自动缩放器（HPA）无法查询CPU/指标，因为metrics-server未部署或运行不正常。

Official Documentation

https://kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/#support-for-metrics-apis

Workarounds

  1. 85% success Deploy metrics-server using the official manifest: `kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml` 
    Deploy metrics-server using the official manifest: `kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml`
  2. 80% success Check metrics-server pod logs: `kubectl logs -n kube-system deployment/metrics-server` and verify it can connect to kubelets. Common fix: add --kubelet-insecure-tls flag if using self-signed certs. 
    Check metrics-server pod logs: `kubectl logs -n kube-system deployment/metrics-server` and verify it can connect to kubelets. Common fix: add --kubelet-insecure-tls flag if using self-signed certs.

中文步骤

  1. Deploy metrics-server using the official manifest: `kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml`
  2. Check metrics-server pod logs: `kubectl logs -n kube-system deployment/metrics-server` and verify it can connect to kubelets. Common fix: add --kubelet-insecure-tls flag if using self-signed certs.

Dead Ends

Common approaches that don't work:

  1. Install a custom Prometheus adapter instead of metrics-server 70% fail

    HPA requires the resource metrics API, which is typically served by metrics-server. Prometheus adapter serves custom metrics, not resource metrics.

  2. Set --horizontal-pod-autoscaler-use-rest-clients=false on kube-controller-manager 90% fail

    This flag is deprecated and removed in recent Kubernetes versions (1.23+). It no longer affects HPA behavior.