README.md

Kubernetes demo with Minikube

---

Requirement

docker has been installed

git command has been installed

Note: This demo only work with machine that running on Linux or Darwin and have arm64 or x86_64 architecture

Assumption

We have 3 application as 3 REST-API:

• demoflaskapi
• demoflaskapi1
• demoflaskapi2

It will be accessed with domain http://demo-service-test.local

We have route following:

http://demo-service-test.local                 -> demoflaskapi
http://demo-service-test.local/demoflaskapi1   -> demoflaskapi1
http://demo-service-test.local/demoflaskapi2   -> demoflaskapi2

This demo following this design:

Usage

Prepare ingredients

Firstly, you must run make install to add execute permission for install script. After that you can install minikube by :

make install-minikube # for install minikube to run k8s in local machine

Secondly you must export some environment variable that is authentication information to push local image to Docker Registry:

export DOCKER_HUB_USER=<username> # username of docker hub or private registry
export DOCKER_HUB_PASSWORD=<password> # password of docker hub or private registry
export DOCKER_HUB_REPOSITORY=<registry-url> # username of docker hub or private registry url

To build image and push to registry use:

make build-image

After script execution successfully you can see:

Please use yellow string (image tag be pushed to registry) to replace tag name and replace registry url in application manifests:

./manifest/application/demoflaskapi.yaml
./manifest/application/demoflaskapi1.yaml
./manifest/application/demoflaskapi2.yaml

It will look like:

Cooking

Firstly, make sure minikube has been started.

Secondly, make sure kubectl installed, if it were not installed, run command minikube kubectl

Thirdly, create some namespace to deploy application by kubectl:

kubectl create namespace demo-test    #namespace for deploy demo application
kubectl create ingress-nginx          #namespace for deploy ingress controller to handle http request and forward to service

Deploy metrics server

The metrics-server fetches resource metrics from the kubelets and exposes them in the Kubernetes API server through the Metrics API for use by the HPA and VPA To deploy metrics-server, please use command following:

kubectl apply -f manifest/metrics

To make sure ingress controller was deployed successfully use command kubectl top node

If output look like image bellow. Congratulation!

Deploy Nginx-Ingress-Controller

An ingress controller acts as a reverse proxy and load balancer. It implements a Kubernetes Ingress. The ingress controller adds a layer of abstraction to traffic routing, accepting traffic from outside the Kubernetes platform and load balancing it to Pods running inside the platform

To deploy nginx ingress controller, please use command following:

kubectl apply -f manifest/ingress-controller

To make sure ingress controller was deployed successfully use command kubectl get pod -n ingress-nginx

If output look like image bellow. Congratulation!

Deploy our application

In this demo we have 3 application that were defined by 3 manifest file in manifest/application

we will use kubectl apply -f for each file to deploy application

kubectl apply -f manifest/application

To make sure our application was deployed successfully use command kubectl get pod -n demo-test

If output look like image below. Congratulation!

Check ingress domain deployed by kubectl get ingress -n demo-test Output:

Add domain to host files in /etc/hosts:

127.0.0.1     demo-service-test.local 

Use minikube service expose port to forward request from outside to ingress-nginx-controller:

minikube service ingress-nginx-controller -n ingress-nginx --url

Output look like:

Finally, use port in first line to access to application look like:

Check horizontal pod autoscaling

Use kubectl get hpa -n demo-test

Output look like: