oam/knowledge base/drain a k8s cluster node.md

Drain a K8S cluster node

1. mark the node as unschedulable (cordon):

   $ kubectl cordon kworker-rj2
   node/kworker-rj2 cordoned
   

2. remove pods running on the node:

   $ kubectl drain kworker-rj2 --grace-period=300 --ignore-daemonsets=true
   node/kworker-rj2 already cordoned
   WARNING: ignoring DaemonSet-managed Pods: kube-system/calico-node-fl8dl, kube-system/kube-proxy-95vdf
   evicting pod default/my-dep-557548758d-d2pmd
   pod/my-dep-557548758d-d2pmd evicted
   node/kworker-rj2 evicted
   

3. do to the node what you need to do

4. make the node available again:

   $ kubectl uncordon kworker-rj2
   node/kworker-rj2 uncordoned
   

Sources

• How to drain a node in Kubernetes