Do kubernetes master and Worker nodes need standalone machines?

Question

Im starting to work on Kubernetes and there is one thing that is bugging me. The question might be silly, but Im unable to find an answer for it.

In Kubernetes, we have Master node(Control Plane) and the Worker nodes(where the containers are hosted).

Now, when you say master node, is it a physical or virtual server where you install the required applications/master node processes like API server, scheduler or etcd?

What I mean is, is master node on a windows/linux machine or server?

And also, same goes for Worker nodes.

Thanks!

Answer 1

For production scenarios it is generally recommended to host the control plane components on nodes separate from the application workloads. However it is possible and there exists perfectly valid scenarios (like local development environments) where Kubernetes runs on a single node, mixing control plane components with application workloads

A Kubernetes cluster consists of a set of worker machines, called nodes, that run containerized applications. Every cluster has at least one node (otherwise it would not be a cluster). The nodes can be either virtual or physical.

In Kubernetes you further distinguishes between control plane components and application workloads.

The control plane's components make global decisions about the cluster (for example, scheduling), as well as detecting and responding to cluster events.

Control plane components can be run on any machine in the cluster, however the general convention is to run the control plane across a set of machines (physical or virtual) that do not run any application workloads (they will instead run on worker nodes).

Please see the Kubernetes Components documentation for more details