Kubernetes architecture
As you already know, Kubernetes is used to orchestrate fleets of containers that run on multiple servers that make up a Kubernetes cluster. Those servers are often called nodes, and nodes can be virtual machines running on-premises, in the cloud, or bare-metal servers. You can even combine different nodes in one Kubernetes cluster (for example, several nodes represented by VMs plus a few others as bare-metal servers).
There are two distinguished node types in Kubernetes:
- Control plane nodes (sometimes also called master nodes)
- Worker nodes
It is the worker nodes where the containerized applications run, and it is the control plane nodes where the K8s cluster management components run. We can see this in more detail in Figure 5.1.
Figure 5.1 – Kubernetes components
Control plane nodes run several specialized K8s services and make global decisions about the Kubernetes cluster, such as scheduling containerized...