Kubernetes is an open source container orchestration tool developed by Google. You can run and manage container-based workloads by using Kubernetes. The most common Kubernetes use case is to deploy an array of interconnected microservices, building an application in a cloud native way. You can create Kubernetes clusters that can span hosts across on-premise, public, private, or hybrid clouds.

Traditionally, applications were deployed on top of a single operating system. With virtualization, you can split the physical host into several virtual hosts. Working on virtual instances on shared resources is not optimal for efficiency and scalability. Because a virtual machine (VM) consumes as many resources as a physical machine, providing resources to a VM such as CPU, RAM, and storage can be expensive. Also, you might see your application degrading in performance due to virtual instance usage on shared resources.

247 OpenShift Kubernetes Overview
Figure 1. Evolution of container technologies for classical deployments

To solve this problem, you can use containerization technologies that segregate applications in a containerized environment. Similar to a VM, a container has its own filesystem, vCPU, memory, process space, dependencies, and more. Containers are decoupled from the underlying infrastructure, and are portable across clouds and OS distributions. Containers are inherently much lighter than a fully-featured OS, and are lightweight isolated processes that run on the operating system kernel. VMs are slower to boot, and are an abstraction of physical hardware. VMs run on a single machine with the help of a hypervisor.

You can perform the following actions by using Kubernetes:

  • Sharing resources

  • Orchestrating containers across multiple hosts

  • Installing new hardware configurations

  • Running health checks and self-healing applications

  • Scaling containerized applications

Kubernetes components

Table 1. Kubernetes components
Component Purpose


Runs on every node in the cluster and maintains the network traffic between the Kubernetes resources.


Governs the state of the cluster.


Allocates pods to nodes.


Stores cluster data.


Validates and configures data for the API objects.


Runs on nodes and reads the container manifests. Ensures that the defined containers have started and are running.


Allows you to define how you want to run workloads. Use the kubectl command to interact with the kube-apiserver.


Node is a physical machine or a VM in a Kubernetes cluster. The control plane manages every node and schedules pods across the nodes in the Kubernetes cluster.

container runtime

container runtime runs containers on a host operating system. You must install a container runtime on each node so that pods can run on the node.

Persistent storage

Stores the data even after the device is shut down. Kubernetes uses persistent volumes to store the application data.


Stores and accesses the container images.


The pod is the smallest logical unit in Kubernetes. A pod contains one or more containers to run in a worker node.

Kubernetes resources

A custom resource is an extension of the Kubernetes API. You can customize Kubernetes clusters by using custom resources. Operators are software extensions which manage applications and their components with the help of custom resources. Kubernetes uses a declarative model when you want a fixed desired result while dealing with cluster resources. By using Operators, Kubernetes defines its states in a declarative way. You can modify the Kubernetes cluster resources by using imperative commands. An Operator acts as a control loop which continuously compares the desired state of resources with the actual state of resources and puts actions in place to bring reality in line with the desired state.