Components

Introduction

A Kubernetes consists of a single or set of compute machines and is referred to as node(s).

When these nodes are installed with Kubernetes and are interconnected with each other in the same isolated space, then it becomes a cluster. A cluster must have at least one node.

A node consists of two types, and they are:

• worker node(s)
• master node(s)

A Kubernetes comes with two components, and they are:

• Control plane components which are installed on master node(s), and
• Data plane components which are installed on worker node(s)

Control Plane Components

It consists of collection processes that are responsible for the global decisions of the cluster as well as detecting and responding to cluster events.

Here are the processes that make up the control plane components and what they are responsible for:

kube-apiserver

• This component exposes the APIs of Kubernetes
• It is serves as the frontend for the control plane
• It is designed to scale horizontally and balance traffic between those scaled instances of the component

kube-scheduler

• It watches for newly created Pods that have not yet been assigned to a node and selects a node for it to run.
• There are several factors taken into consideration by this component, and they are:
  • Individual and collective resource requirements
  • Hardware, software or policy constraints
  • Affinity and anti-affinity configurations
  • Data locality
  • Deadlines

kube-controller-manager

• It runs controller processes

• There are various types of controllers and some of them are: -Node controller: Responsible for observing the nodes going down and triggering appropriate actions. -Job controller: Responsible for observing Job objects and creating Pods to run those jobs to completion. -ServiceAccount controller: Responsible for creating the default ServiceAccount for new namespaces. -EndpointSlice controller: Responsible for populating EndpointSlice objects to provide links between Services and Pods.

  Note

  Each controller is a separate process; however, they are compiled into a single binary and run as a single process.

etcd

A highly available and consistent key-value store is used to persist the Kubernetes cluster data.

cloud-controller-manager

• It is a component that allows you to link your cluster with your cloud service provider's API
• It's only available if your cluster is created by one of the cloud service provider
• It is introduced to separate interaction with the cloud platform from interaction that is specific to cluster

• There are various types of controllers, and some of them are:

  • Node controller: Responsible for verifying with the cloud service provider the state of the node when it is unresponsive
  • Route controller: Responsible for setting up routes in the underlaying cloud infrastructure
  • Service controller: Responsible for provisioning and de-provisioning cloud load balancers
  Note

  Each controller is a separate process; however, they are compiled into a single binary and run as a single process.

Data Plane Components

The data plane components run on every worker node.

kubelet

• It is responsible for ensuring that containers are running in Pod
• It does it by reading the containers described in PodSpecs
• It does not manage containers that are not created by Kubernetes

kube-proxy

• It is a network proxy responsible for implementing certain parts of the Service object
• It maintains the network rules on worker nodes, which allows network communication to Pods
• It uses the packet filtering layer of the operating system if it exists and is available to use; otherwise, it forwards the traffic by itself

Container Runtime

• It is responsible for managing the execution and lifecycle of containers within the Kubernetes
• There are various container runtime, and they are:
  • containerd
  • cri-o
  • Any container runtime that adheres to the Container Runtime Interface