This is a very simple demo application to demonstrate container technologies and container orchestration. Hopefully by the end of this we will all learn some useful basics.
Docker is almost synonymous with container technologies and for good reason. They were one of the first in the industry to make it easy for developers to create containers.
Kubernetes is a container orchestrator. In simple terms what it means is that it makes sure containers run in the given infrastructure. It does much more than that but for our use case this is enough.
Full disclaimer I am not a python developer. But when I thought about creating this demo I thought what language would be most accessible to most attendees. Python was the clear winner.
The application is quite simple. I have a list of quotes. Everytime you visit the page at path /hello/{name} it will show you one quote and its author.
To create the docker container for this application we have the Dockerfile. The content of the file is as follows.
FROM python:alpine3.7
WORKDIR /app
COPY requirements.txt /app
RUN pip install --trusted-host pypi.python.org -r requirements.txt
COPY . /app
EXPOSE 80
ENTRYPOINT [ "python" ]
CMD [ "./hello.py" ]
This is pretty straight forward as dockerfiles go.
docker build -t <username>/<imagename>:<tag> .
docker login
docker push <username>/<imagename>:<tag>
Public dockerhub is great for learning about docker images and getting other people to easily use and access your application. But when you are writing software you don't want just any one to access, you would want to use a private registry.
IBM Cloud gives you a free image registry.
Go to cloud shell
https://cloud.ibm.com/shell
If it asks for a key use ikslab.
Login to your IBM Cloud account. In your dashboard you should see a cluster.
ibmcloud ks clusters
This would list all the clusters. You should see only one.
ibmcloud ks cluster config -c $CLUSTER-NAME
This will print a export KUBECONFIG=... command. Copy and paste that in your terminal.
This sets the kubectl with the right config.
Apply the deployment
kubectl apply -f k8s/deployment.yaml
Apply the service
kubectl apply -f k8s/service.yaml