Contanerizing a pre-trained, sklearn model that has been trained to predict housing prices in Boston according to several features, such as average rooms in a home and data about highway access, teacher-to-pupil ratios, and so on. Data was initially taken from Kaggle,the data source site. This project tests our ability to operationalize a Python flask app—in a provided file, app.py—that serves out predictions (inference) about housing prices through API calls. This project could be extended to any pre-trained machine learning model, such as those for image recognition and data labeling. The app is containerized via docker desktop and kubernetes cluster. It is verified by circleci.

1. cloned devops udacity repo (starter code)
2. created .circleci/config.yml
3. set up virtual environment - python3 -m venv ~/.devops - source ~/.devops/bin/activate
4. installed dependencies via makefile - make install
5. linting project (hadolint for Dockerfile / pylint for Python) -make lint
6. Completed steps in Dockerfile
7. Completed run_docker.sh
8. Running run_docker.sh - ./run_docker.sh
9. Running predictions via make_predictons.sh in another terminal. - ./make_predictions.sh
10. Added log statement in app.py
11. Repeat steps 8 and 9
12. Added the logs in docker_out.txt
13. Completed upload_docker.sh
14. Uploaded image on dockerhub via upload_docker.sh - ./upload_docker.sh
15. Configure Kubernetes to run locally - kubectl config view
16. Completed run_kubernetes.sh
17. Running run_kubernetes.sh - ./run_kubernetes.sh
18. checking status of kubernetes pod - kubectl get pod
19. Repeat step 17 once status is running.
20. Making a Prediction via make_predictions.sh - ./make_prediction.sh
21. Add the logs in kubernetes_out.txt
22. Delete the kubernetes cluster
23. Pushed project directory to github
24. Set up Circleci project
25. Added a status badge.

1. .circleci
   • .circleci/config.yml for circleci configuration
2. model_data
   • contains data for machine learning app.
3. output_txt_files
   • docker_out.txt
     • contains log info of prediction in docker container. -kubernetes_out.txt
     • contains log info of prediction in kubernetes cluster.
4. app.py -application source code
5. Dockerfile
   • configuration for Docker image
6. make_prediction.sh
   • bash file to get prediction from app
7. Makefile
   • commands to setup environment and linting
8. requirements.txt
   • contains dependencies
9. run_docker.sh
   • bash file to build and run docker image.
10. run_kubernetes.sh
    • bash file to build and run kubernetes cluster.
11. upload_docker.sh
    • bash file to upload docker to dockerhub.
12. README.md
    • Project summary, steps and repository overview