Kubernetes The Hard Way With Ansible Save Abandoned

Introduction

This is an attempt to automate the creation of k8s cluster proposed in the tutorial kubernetes-the-hard-way.

Cluster details

• k8s version: 1.10.2
• Container runtime: Docker
• Network plugin: Flannel
• DNS plugin: CoreDNS
• gVisor supported.

Prerequisites

Before you begin, you need to install all these tools:

• Ansible
• Vagrant
• Virtualbox
• kubectl

You need also to have at least 4 GB of free memory.

How to use

Create the VMs with Vagrant

$ vagrant up

Deploy k8s cluster

$ ansible-playbook k8s.yml

Verification

Check the k8s master's components

$ kubectl get componentstatus
NAME                 STATUS    MESSAGE              ERROR
controller-manager   Healthy   ok                   
scheduler            Healthy   ok                   
etcd-2               Healthy   {"health": "true"}   
etcd-0               Healthy   {"health": "true"}   
etcd-1               Healthy   {"health": "true"}   

Check the k8s workers

$ kubectl get nodes
NAME          STATUS    ROLES     AGE       VERSION
k8s-worker1   Ready     <none>    19s       v1.10.2
k8s-worker2   Ready     <none>    20s       v1.10.2
k8s-worker3   Ready     <none>    20s       v1.10.2

Check DNS add-on

$ kubectl get pod -l k8s-app=coredns -n kube-system
NAME                      READY     STATUS    RESTARTS   AGE
coredns-b8d4b46c8-t2zpf   1/1       Running   0          37m

Check Dashboard

$ kubectl get pod -l k8s-app=kubernetes-dashboard -n kube-system
NAME                                    READY     STATUS    RESTARTS   AGE
kubernetes-dashboard-7c5d596d8c-vp85k   1/1       Running   0          8m

Accessing Dashboard

Get the port on which the dashbord is exposed

$ kubectl -n kube-system get service kubernetes-dashboard
NAME                   TYPE       CLUSTER-IP   EXTERNAL-IP   PORT(S)         AGE
kubernetes-dashboard   NodePort   10.32.0.63   <none>        443:30585/TCP   23m

The Dashbord is accessible at https://<Worker-IP>:30585

To log in, select Token and insert the token given by the following command:

$ kubectl describe serviceaccount kubernetes-dashboard -n kube-system | grep Tokens | awk '{print $2}' | xargs kubectl -n kube-system describe secret  | grep '^token' | awk -F ':' '{print $2}'

Check the monitoring pods

$ kubectl get pods  -l task=monitoring --all-namespaces
NAMESPACE     NAME                                   READY     STATUS    RESTARTS   AGE
kube-system   heapster-69b5d4974d-kxgjw              1/1       Running   0          2m
kube-system   monitoring-grafana-69df66f668-m59hp    1/1       Running   0          2m
kube-system   monitoring-influxdb-78d4c6f5b6-gh56w   1/1       Running   0          2m

Access Grafana

Get the port on which grafana is exposed:

$ kubectl -n kube-system get service monitoring-grafana  
NAME                 TYPE       CLUSTER-IP    EXTERNAL-IP   PORT(S)        AGE
monitoring-grafana   NodePort   10.32.0.238   <none>        80:31244/TCP   7m

the dashboard is accessible at http://<Worker-IP>:31244

License

The source code in this repo is licenced under the GPL 3