Skip to content

Latest commit

 

History

History
92 lines (54 loc) · 3.59 KB

Chapter3.md

File metadata and controls

92 lines (54 loc) · 3.59 KB

Chapter 3. Deploying a Kubernetes Cluster

After building the applications in containers, it needs to deploy into a reliable, scalable distributed system(K8s cluster). For this, there are many services provided by cloud which are quite easy to get started but sometimes it could be costly. Minikube(single-node cluster) tool provides a way to run cluster on local computer or VM. Kubeadm tool helps in running k8s on bare metal(cluster of Raspberry Pi single board computers)

Installing Kubernetes on a Public Cloud Provider

In Google Container Service:

gcloud config set compute/zone asia-south1-b # default zone

gcloud container clusters create kuar-cluster # cluster creation

gcloud auth applicaton-default login # credentials

In Azure Container Service:

az group create --name=kuar --location=westus # resource-group creation

az acs create --orchestrator-type=kubernetes --resource-group=kuar --name=kuar-cluster # cluster creation

az acs kubernetes get-credentials --resource-group=kuar --name=kuar-cluster # credentials

az acs kubernetes install-cli # installing kubectl

In AWS(Amazon Web Services):

Till the release of book, AWS does not offer any hosted Kubernetes services, but there are many ways to do that. But now there is on Amazon Elastic Container Service for Kubernetes which does the same.

Check out every solution here: Pick Right Solution

Installing Kubernetes Locally Using Minikube

For local development(single-node cluster), learning, and experimentation.

minikube start # starting the cluster locally, creates a local VM, creates a local kubectl config

minikube stop

minikube delete

There are other options too to run Kubernetes on Raspberry Pi if wanted to run K8s on bare metal cluster without paying a lot.

The Kubernetes Client(Kubectl)

A command line tool to interact with Kubernetes API and used to manage most K8s objects such as pods, ReplicaSets, and services as well as used for exploration and verification of cluster.

Some commands for usage of Kubectl on the running cluster(most of commands are self-explanatory):

kubectl version

kubectl get componentstatuses # controller-manager regulates controller behavior; scheduler used for placing pods onto different nodes; etcd is the API object storage for the cluster

kubectl get nodes # listing nodes in the cluster

kubectl describe nodes <node name>

Cluster Components

Components run in the kube-system namespace(folder in a filesystem)

  • Kubernetes Proxy: It is responsible for routing network traffic to load-balanced services in the cluster. K8s uses DaemonSet which helps in achieving proxies on every node of the cluster.
kubectl get daemonSets --namespace=kube-system kube-proxy
  • Kubernetes DNS: K8s runs a DNS server for providing names and discovery of the services that are defined in the cluster. It runs as a Deployment and service in K8s for managing replicas and load-balancing.
kubectl get deployments --namespace=kube-system kube-dns

kubectl get services --namespace=kube-system kube-dns
  • Kubernetes UI: Kubernetes also consists of UI dashboard(a GUI).
kubectl get deployments --namespace=kube-system kubernetes-dashboard

kubectl get services --namespace=kube-system kubernetes-dashboard

kubectl proxy # starts up a server on localhost:8001

This chapter was mostly about the components and objects kubernetes provides.