How shivanshvermaa/mlops-deployment works
Main Components
• FastAPI Application: Represented by the “app” folder (with main.py, logging configuration, and model scripts). This is the core API server that handles requests, serves the ML model, and logs events.
• Logging & Monitoring: Indicate two parts – internal application logs (app/logs) and external fastapi-logs that are processed by Filebeat. The “filebeat” folder contains container configuration, showing that logs are being collected and forwarded (presumably to an ELK stack).
• Containerization & Orchestration: Show the Dockerfile and docker-compose.yaml, indicating that the application and its associated services are packaged into containers. Also include the Heroku deployment file (heroku.yml) since it represents another deployment target.
Relationships and Interactions
• Incoming User Traffic flowing to the FastAPI container running on Docker within the AWS EC2 instance.
• If applicable, add interactions between the FastAPI application and AWS services like S3 (for accessing or storing objects) and possibly other AWS resources for scaling or security.
Important Architectural Patterns & Technologies
• Microservice/Containerized Architecture: Emphasize that the application is containerized (using Docker and docker-compose) and modularized into separate components (API, logging, CI/CD).
• CI/CD Integration: Note that GitHub Actions provide automated builds and deployments, orchestrating changes from code commit to production in AWS.
• Observability: Mark the logging components (Filebeat and log directories) which interface with the ELK stack for centralized logging and monitoring.
• Cloud Deployment: Indicate that the application is designed to be deployed on AWS EC2 with associated configurations for security, scaling, and network access.