kubectl uncordon removes a node’s unschedulable status, allowing newly scheduled pods to use the node again. It does not evict pods already running there. section 5.2 section 5.2.1
Use uncordon as the final step of node maintenance:
- Cordon the node to prevent new pods from being scheduled onto it. section 5.2.1
- Drain the node to evict eligible pods and let controllers recreate them elsewhere. section 5.2.1
- Run kubectl uncordon <node-name> to re-enable scheduling after maintenance. section 5.2
- Verify the node no longer shows SchedulingDisabled with kubectl get no. section 5.2
Restore scheduling after draining a worker
After draining kind-worker and confirming workloads have moved, run kubectl uncordon kind-worker. Kubernetes reports node/kind-worker uncordoned, and new pods can again be scheduled there. section 5.2
