End-to-End Service Setup

A. Set Up a Conda Environment

1. Follow basic conda commands for environment management here.

2. Create and activate a new conda environment:

   conda create -n end-to-end python=3.9
   conda activate end-to-end

B. Write Basic Functionality

1. Always use functions for repetitive tasks.

2. Create two scripts:

   • main.py: Holds the main functionality and calls the necessary functions from utility_v1.py.
   • utility_v1.py: Contains reusable functions.

C. Create the API

1. Install required libraries:

   pip install fastapi uvicorn

2. Create an API script with a POST endpoint and functionality to process input and return a JSON response.

3. Example of API startup:

   uvicorn api_v1:app --reload

   Or, include the following in your script:

   if __name__ == "__main__":
       uvicorn.run(app, host="0.0.0.0", port=5001)

   Run the script:

   python api_v1.py

4. Test the API:

   Using curl:

   curl -X 'POST' 'http://127.0.0.1:5001/echo/' -H 'Content-Type: application/json' -d '{"message": "Hello World"}'

   Or use Postman.

D. Dockerization

1. Export the environment libraries:

   conda list --export > requirements.txt

2. Clean unnecessary library versions in requirements.txt (recommended).

3. Dockerize the application:

   • Dockerfile: Copy necessary files and set up the environment.
   • docker-compose.yml: Automate Docker setup and start the service.

4. Build and run the service:

   docker-compose build
   docker-compose up -d

   Or combine the steps:

   docker-compose up -d --build

5. Check container logs:

   docker logs <container_id>

   Example:

   docker logs 07c91a7f6c40

6. Test using curl or Postman as mentioned in step C.

E. Access and Interact with the Docker Container

1. List running containers:

   docker ps

2. Access the container shell:

   docker exec -it <container_id> /bin/bash

   Example:

   docker exec -it 07c91a7f6c40 /bin/bash

3. Run Python scripts inside the container:

   python main.py

4. Run one-time commands without entering the container shell:

   docker exec <container_id> python main.py