Скачать с ютуб [02.01.01] Draining a Worker Node during the Kubernetes Cluster Upgrade в хорошем качестве

[02.01.01] Draining a Worker Node during the Kubernetes Cluster Upgrade

In this video I am highlighting the drainage process of a worker node. In my previous video [02.01] I failed to execute that step which caused a failure of my database cluster. However, as I followed additional steps that were highlighted in some of the outputs, I was able to successfully drain the node during the upgrade of the worker node to version 1.31.4 and successfully have the database pod, which was evicted from this node during the process, restart. In this video I am showcasing these additional steps which are crucial to prevent additional troubleshooting steps and make the upgrade process much smoother. 0:00 Introduction & Purpose for this video 0:55 Setup 2:00 Identify pods to watch during the upgrade process 2:46 Initiate update process 3:22 Updating kubeadm 4:50 draining Kubernetes worker 6:20 Updating kubectl and kubelet 8:15 Reintegrating Kubernetes Worker Node into the Cluster 10:17 Summary Sources: Kubernetes documentation - kubernetes.io Kubernetes upgrade path - https://v1-31.docs.kubernetes.io/docs... Commands from the video: On the Worker 3:24 sudo apt-mark unhold kubeadm 3:34 sudo apt-cache madison kubeadm 3:49 sudo apt install -y kubeadm='1.31.4*' 4:06 sudo apt-mark hold kubeadm 4:27 sudo kubeadm upgrade node On the MASTER 5:08 kubectl drain k8s-worker1 --ignore-daemonsets 5:52 kubectl drain k8s-worker1 --ignore-daemonsets --delete-emptydir-data On the Worker 6:34 sudo apt-mark unhold kubelet kubectl 6:58 sudo apt install -y kubelet='1.29.12-*' kubectl='1.29.12-*' 7:33 sudo apt-mark hold kubelet kubectl 7:49 sudo systemctl daemon-reload 7:58 sudo systemctl restart kubelet From the MASTER 8:15 kubectl uncordon k8s-worker1