Kubernetes

K8S on Minikube

I have setup my Kubernetes environment locally. For this setup i am using a HP Workstation as this is the best option i have to do something local. To save on costs running this in the cloud i bought a secondhand machine

I do intend to also rollout images on AWS but that will follow in a later stage

Taking my notes here, based on information from Kubernetes.io The setup is described for a Linux environment, ofcourse. The instruction apply to MacOs as well, just download the installers for OSX.

Pre check and dependencies

First you need to be sure your machine is supporting Virtualization, otherwise all effort is lost.
VMX or SVM should be highlighted, provided you run Linux.
grep -E --color 'vmx|svm' /proc/cpuinfo

Virtualbox

Next, make sure Virtualbox is installed, find it here for your OS: https://www.virtualbox.org/wiki/Downloads.
First add the repo to your sources.list and import the key

Add the following line to /etc/apt/sources.list
deb http://ftp.debian.org/debian stretch-backports main contrib

Download the key and import it
wget -q https://www.virtualbox.org/download/oracle_vbox_2016.asc -O- | apt-key add -

Update apt and next install virtualbox
apt install virtualbox-5.2

Kubectl

Get the installer and make it executable, commands a run as root user

LINUX
curl -LO https://storage.googleapis.com/kubernetes-release/release/`curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt`/bin/linux/amd64/kubectl && chmod +x kubectl

OSX
curl -LO "https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/darwin/amd64/kubectl"

Move it to the right dir in your path
mv kubectl /usr/local/bin

Minikube

Do the same thing for minikube

LINUX
curl -Lo minikube https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64 && chmod +x minikube

OSX
curl -Lo minikube https://storage.googleapis.com/minikube/releases/latest/minikube-darwin-amd64 && chmod +x minikube

Now move it into your path
mv minikube /usr/local/bin

Check Setup

Now switch back to your login user and try if things are working

This will download some images and install them using Virtualbox.

Start Minikube
minikube start

Check the status
minikube status

which will provide you with info:

minikube
type: Control Plane
host: Running
kubelet: Running
apiserver: Running
kubeconfig: Configured

Check if kubectl can access the cluster
kubectl cluster-info

Kubernetes master is running at https://192.168.99.100:8443
KubeDNS is running at https://192.168.99.100:8443/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy

To further debug and diagnose cluster problems, use 'kubectl cluster-info dump'

So, you are good to go from here and make a deployment

Hello World

Just a small hello world kind of test and exposing to local port 8888

Create the deployment

Create the deployment of hello-minikube
kubectl create deployment hello-minikube --image=k8s.gcr.io/echoserver:1.10

Expose the deployment

Expose the deployment to local port 8888
kubectl expose deployment hello-minikube --type=NodePort --port=8888

Check Status

Now, we can check for the deployment which is a pod
kubectl get pod

This will produce you with some info:

NAME                              READY   STATUS              RESTARTS   AGE
hello-minikube-64b64df8c9-krr6p   0/1     ContainerCreating   0          37s

To see where the service is running:
minikube service hello-minikube --url

This shows you the url:

http://192.168.99.100:32463

End of the show

When you have stopped enjoying the magic:
minikube stop

To delete all the good work

Removing just the deployment of hello-minikube
kubectl delete services hello-minikube

To remove the minikube cluster
minikube delete
This completes the instruction to setup a local test environment. I will continue to add examples and tips&tricks soon.

Scaling Applications

Jumping ahead a little bit now to scaling of applications in Kubernetes.

- Scaling in Kubernetes is done using the Replication Controller.
- The replication controller will ensure a number of replicas is running at all times.
- Pods created will automatically be replaced upon failure, termination or delete.
- It can also be used to make sure a pod will be running even after a reboot.

Configuration Example

This is done by changing your configuration a little. Here we will run 2 replicas of the demo app i created preliminary and have put on docker hub.

apiVersion: v1
kind: Replicationcontroller
metadata:
 name: demo
spec:
 replicas: 2
 selector:
  app:demo
 template:
  metadata:
  labels:
   app: demo
 spec:
  containers:
  - name: demo
    image: phenixops/demo
    ports:
    - containerPort: 80