Kubernetes Components

• Refer Here for kubernetes master and node components
• K8s cluster is collection of nodes
• Nodes are classified as
  • Master node
  • nodes (Workers)
• K8s is controlled by master nodes so the components in master are referred as control plane components
• Control plane (master) components
  1. kube-apiserver:
     • Front end of k8s
     • All communications with in cluster
  2. etcd:
     • Highly available key-value store
     • memory of k8s cluster
  3. kube-scheduler:
     • Scheduling newly created pods on the nodes
  4. kube-controller-manager:
     • Responsible for maintaining desired state
     • Kube-controller manager has the following controllers
       • Node controller: Responsible for noticing and responding when node goes down
       • Replication Controller: Responsible for mainting the correct number of pods (desired state)
       • Endpoint controller: Populating the Endpoints object
  5. cloud-controller manager:
     • This component embeds cloud-specific control login.
• Node components

  1. kubelet:

     • agent that runs on every node in cluster.
     • It recieves instructions from k8s control plane and will do the job

  2. kube-proxy

     • it is a network proxy that runs on each node in k8s cluster
     • It maintains network rules on the node. The rules allow network communications to your pods

  3. Container runtime

     • Container runtime is software responsible for running containers
     • k8s supports Docker, containerd, CRI-o runtime or any implementation of kubernetes CRI.
• Addons:
  1. DNS
  2. Web UI
  3. Container Resource Monitoring
  4. Cluster Level Logging

K8s Setup

• K8s has many flavours/installation types

  • Desktop => minikube
  • Production => Refer Here

• Since there are many ways of installing k8s, we will be using

  1. kubeadm
  2. EKS (Elastic Kubernetes Services)
  3. AKS (Azure Kubernetes Services)

Using Kubeadm to setup k8s cluster (Single Master)

• We will be using 3 Ubuntu 18 Linux VMS Refer Here for installation steps
• For prereq’s Refer Here
• create three vms in aws/azure with 2vpu and 2 GB of RAM at minimum
• Install Docker Runtime on all nodes by following steps from here
• Now we need to install kubeadm, kubectl Refer Here on all nodes

Next steps

• Now lets create the k8s cluster using kubeadm Refer Here