Today I registered my user name patclaffey on the Docker Hub web-site - I am now the proud owner of a personal patclaffey Docker Hub account. The first thing I did on Docker Hub was create a public repository for my Jupyter / Pandas image. I navigated to the Dashboard tab, pressed the Create Repository + button and followed its dialog:

1. Choose a namespace (Required) - defaults to my user name patclaffey
2. Choose a name (Required) - I choose jupyter_pandas as my repository name
3. Add a short description (Required) - I added a short description
4. Add markdown to the full description field - I left this blank initially, and filled this field in later
5. Set it to be a private or public repository - I left this as public, the default value.

I now have the empty Docker Hub repository for my project. This is a repository for project Docker images. To add an image I need to leave Docker Hub and return to the docker tool on my host laptop.

To use Docker Hub, the Images on the host must follow the Docker Hub naming convention. This naming convention is user_name/reporitory_name:tag. Here is the command I used to prepare the image prior to pushing it to Docker Hub:

docker build - t "patclaffey/jupyter_pandas:v2" 

The docker push command is used to push the image to Docker Hub:

docker push patclaffey/jupyter_pandas:v2

I navigated back to the Docker Hub dashboard on my browser. I verified that my image was successfully loaded by going to the Tags sub-tab. I verified that the image tagged v2 was available

Benefits of Docker Hub

The major benefit of Docker Hub is that my image is now publically available to any user from any Docker host using the following docker pull command:

docker pull patclaffey/jupyter_pandas:v2

A more intangible benefit of Docker Hub is that by making the image public, it focuses the user on the quality of the image. So that will be the subject of a follow on blog. I would like to apply the Docker Hub image best practices to create the highest quality possible image for Jupyter and Pandas.