docker-compose.yaml

version: '3.8'  # Specify docker-compose version

services:
  mlflow:
    image: ghcr.io/mlflow/mlflow:latest  # Use the official MLflow Docker image
    container_name: cords-mlflow
    ports:
      - 4000:4000  # Expose MLflow UI on port 5000
    volumes:
      - ./mlruns:/mlruns  # Mount local directory to store MLflow artifacts and metadatas
    environment:
      - MLFLOW_TRACKING_URI=http://localhost:4000  # Set the tracking URI to the running container
      - MLFLOW_S3_ENDPOINT_URL=  # Optional: If using S3 storage, provide the endpoint URL
      - AWS_ACCESS_KEY_ID=  # Optional: AWS access key (if using S3)
      - AWS_SECRET_ACCESS_KEY=  # Optional: AWS secret key (if using S3)

    command: mlflow server --host="0.0.0.0" --port="4000"
    networks:
      - local

  cords-resource-manager:
    image: cords_resource_manager:latest
    container_name: resource_manager
    ports:
      - "5000:5000"
    volumes:
      - ./policies/:/app/policies
    environment:
      FLASK_APP: app.py
    networks:
      - local
      
networks:
  local:
    driver: bridge