Exercises

  1. Deploy your image to Docker Hub container registry (needs an account, has free tier limits).

  2. Run a container using the python images from Docker Hub.

  3. Try to minimize the size of the docker images produced.

  4. Add checks for out of bound queries in your recommendation function (e.g., http://localhost/?uid=2000 will give a value error on the server and the browser will show that an internal server error occured).

  5. Add a load balancer to the ECS deployment and study what it does.

  6. Replicate the linked tutorial: AWS and Docker

  7. Use cookie cutter data science repository to learn the best practices while training a machine learning moodel in a container.

  8. Experiment with a different container technology such as https://linuxcontainers.org/.