Chapter 3: Working with Nodes
Everyone familiar with Kubernetes knows that the cluster workload runs in nodes, where all Kubernetes pods get scheduled, deployed, redeployed, and destroyed.
Kubernetes runs the workload by placing containers into pods and then schedules them to run on nodes. A node might be a virtual or physical machine, depending on the cluster setup. Each node has the services necessary to run pods, managed by the Kubernetes control plane.
The main components of the node are as follows:
- kubelet: An agent that registers/deregisters the node with the Kubernetes API.
- Container runtime: This runs containers.
- kube-proxy: Network proxy.
If the Kubernetes cluster supports nodes autoscaling, then nodes can come and go as specified by the autoscaling rules: by setting min and max node counts. If there is not much load running in the cluster, unnecessary nodes will be removed down to the minimum nodes set by the autoscaling rules. And when the load...
