With this tutorial, we'll go deeper into what Docker images are and build our own image. We're going to run it locally and finally deploy it.
To see a list of locally available images, use docker images
The TAG refers to a particular snapshot of the image, IMAGE ID is the unique identifier of that image.
Without a specific version number provided, the client defaults to latest
For example, you can pull a specific version of Ubuntu image
docker pull ubuntu:18.04
There is 2 kind of images
- Base images: With no parents, usually images with an OS
- Child images: built on base images, adding functionality
And both can be Official images, typically one word long. Officially maintained and supported by Docker; or User images, created and shared by users. Typically formatted as user/image-name
Let's create our own first image. We'll do an image that sandboxes a simple Flask app. So, let's clone the repository locally with this: git clone https://github.com/prakhar1989/docker-curriculum.git then cd docker-curriculum/flask-app
The next step is to create an image with this app.
This app is written in python, so we'll use Python 3, but more specifically, we'll use python:3-onbuild version.
The onbuild means that some automation are included. A bit like an install script.
Here, it'll COPY a requirements.txt file, RUN pip install and COPY current directory into /usr/src/app
We need now a Dockerfile!
It's a simple text file that contains a list of commands.
So let's do this!
Starting a blank text page we're going to save to the same folder as the flask app; using the name Dockerfile
We specify our base image
FROM python:3-onbuild
Usually, next step is writing commands of copying files and installing dependencies. But onbuild version already took care of that.
We specify the port needed to be exposed. Our flask app is running on port 5000, so let's do this
EXPOSE 5000
Last thing is to write the command to run the app
CMD ["python", "./app.py"]
Save it and now we're ready to build our image.
Before running the command, make sure to use your Docker Hub username: docker login then when you've completed your credentials: docker build -t <DOCKER_HUB_USERNAME>/<APP_NAME> . [Don't forget the dot at end of line. With a space between the name and the dot]
The -t is used for tag name. The dot is used to give location of the directory containing the Dockerfile.
To publish and deploy our app, we'll use AWS Elastic Beanstalk. First, we must publish our image on a registry that can be accessed by AWS. The easiest one here is Docker Hub. So to publish just do: docker push <DOCKER_HUB_USERNAME>/<APP_NAME>
If it's the first time you push an image, the client will ask you to login. Now you can check your image on Docker Hub.
Anyone using Docker can now use that image like this: docker run -p 8888:5000 <DOCKER_HUB_USERNAME>/<APP_NAME>
Now for AWS, we need an active account. Login to your AWS console, and choose Elastic Beanstalk. Create New Application, give the app a name and a description. In "New Environment", create and choose Web Server Environment Fill the info by choosing a domain. Choose Docker from the predefined platform. And upload your code. We have to modify the Dockerrun.aws.json file by changing it's name. Then we upload The environment is setting up, we have to wait a bit!
That dockerrun.aws.json file is a configuration file specifically tailored for AWS. So now, the app is deployed and ready to use with the URL given by AWS