Containerisation Demo - CA675 DCU Week 6

A small ML program run directly through a container

Step 1: Clone the repo

Clone this repo into a VM with docker installed:

git clone https://github.com/josmas/CA675DCU_containerisation.git

Step 2: Build the image

docker build -t jos/classification:1 .

Note that you can modify the name of the container as you see fit.

You can see your new image with docker images

Step 3: Run the container

docker run --rm --name class_1 jos/classification:1

If you have modified the name of the container, you'll also have to modify the invocation command.

Step 4 (Optional): Run the container iteratively

You can run the container iteratively and connect to it through a shell with:

docker run -it --name class_1 jos/classification:1 bash

You can see your container running if you type docker ps from a second terminal.

If you also map your local folder, you can run whatever python scripts you want from the terminal. Look at the docs for Bind Mounts and Volumes from the Docker site.

Step 5 (Optional): Build a generic Python + scikit-learn container

You can specify an ENTRYPOINT and a default CMD as in Dockerfile.entrypoint. To build this image you can use the following:

docker build -t jos/class_entrypoint:1 -f ./Dockerfile.entrypoint .

If you run it as

docker run --rm --name class_ent_1 jos/class_entrypoint:1

you will see the output for classification_sample.py.

If you run it as

docker run --rm --name class_ent_1 jos/class_entrypoint:1 another_python_sample.py

you will see the output for the second script.

References

The code in this repo is based on the article at: https://towardsdatascience.com/beginners-guide-to-data-science-python-docker-3181fd321a5c

Jos, Oct 2019