A client-server based application to demonstrate deployment and packaging abilities of Kubernetes and Docker combination. Following is the list of tools along with useful documentation:

• Docker - Container runtime for packaging applications.
• Kubernetes - Cluster manager for life cycle management of applications.
• Minikube - Kubernetes flavor for single node developement environment.
• Docker Registry - Docker registry hosting application images.

You will need a kubernetes environment for running the application. Please refer to above documentation for k8s installation.

The server is a docker based nginx server hosting our favourite Hello,World page. The docker image for server is available here.

The client is a cURL based utility that will be used to send periodic HTTP requests to the server. The docker image for client is available here.

Download the code using git:

$ git clone https://github.com/MQasimSarfraz/scalable-web-app.git

All the entities will be deployed via kubectl utility using respective yaml files. All the deployment files are kept in scalable-web-app/kubernetes

Deploy the server using:

$ kubectl create -f kubernetes/deployments/server.yaml

You can verify that the server is running using:

$ kubectl get deployment server
$ kubectl describe deployment server

For the server to be accessible it needs to be exposed as a service using:

$ kubectl create -f kubernetes/services/server.yaml

Client issues periodic HTTP request to server and has following configurable attributes:

• Time between each request (optional)

You can configure the time for HTTP request in client.yaml. The format of the schedule string is explained here. Now deploy client using:

$ kubectl create -f kubernetes/deployments/client.yaml

In order to scale the server, update the replicas in kubernetes/deployments/server.yaml to the required number of server. Once done scale the servers using:

$ kubectl apply -f kubernetes/deployments/server.yaml

The server has two versions available:

• server v1 with nginx 1.10.1
• server v2 with nginx 1.11.1

Both of these images are available at smqasims docker registry. By default server:v1 is deployed.

In order to update the server, update the image in kubernetes/deployments/server.yaml to the the newer version i.e smqasims/server:v2. Once done update the servers using:

$ kubectl apply -f kubernetes/deployments/server.yaml

Images can be uploaded using scalable-web-app/app/build-image.sh script. Specify the required tags of images in the script and run it using:

$ chmod +x build-image.sh
$ ./build-image.sh

Cleanup the application using scalable-web-app/kubernetes/scripts/cleanup.sh script:

$ chmod +x cleanup.sh
$ ./cleanup.sh