NotReady
A node status meaning its Ready condition is false or unknown, so the control plane no longer trusts it to run pods.
In more detail
It is a report from the control plane, not a cause: the kubelet may have stopped, the runtime or network plugin may be broken, or certificates may have expired. Conditions and events narrow it down, and journalctl on the host shows the kubelet's own view. After recovery, prove it with a new pod, not just the Ready status.
Examples
kubectl get node node-ajournalctl -u kubelet -n 5 --no-pagerRelated terms
- node condition — A named fact the kubelet reports about its node, such as Ready, MemoryPressure or DiskPressure, each true or false with a reason.
- node heartbeat — The regular signal a kubelet sends to the API server, by renewing its Lease and updating node status, to show the node is alive.
- kubelet — The agent on every Kubernetes node that starts and watches the node's containers and reports back to the API server.
- journalctl — The command that reads the systemd journal, where services on a Linux host write their logs.
Taught in
- Who Owns The Symptom: Control Plane And Nodes — Kubernetes and Cloud Native Associate (KCNA) Exam Prep
- A Name That Outlives Its Pods: Services And DNS — Kubernetes and Cloud Native Associate (KCNA) Exam Prep
- Worker Nodes: kubelet, Runtime, CNI And kube-proxy — Kubernetes: Production-Grade Container Orchestration
- etcd Snapshots And A Tested Restore — Kubernetes: Production-Grade Container Orchestration
- Troubleshooting: NotReady Nodes And kubelet Logs — Kubernetes: Production-Grade Container Orchestration