mlflow-prod-server

Tested on Windows 11

Clone and Run mlflow ready server with minio s3 artifact storage and postgresql persist storage.

The mlflow UI is accessible on http://localhost:3001 and the minio ui http://localhost:9001

Quick Start

git clone https://github.com/4ug-aug/mlflow-prod-server.git

Enter directory

cd mlflow-prod-server

Compose docker containers

docker compose up -d --build

Access the minio storage on port 9001 by ip of server running docker (possibly localhost if you are running locally). Example: http://localhost:9001

Login with: Username: miniouser Password miniospw

Create an access key and paste these into the docker compose file:

[...]

  mlflow:
    restart: always
    build: 
      context: .
      dockerfile: Dockerfile
    container_name: mlflow
    networks:
      - frontend
      - backend
    ports:
      - 3001:3001
    environment:
      - MLFLOW_TRACKING_URI=http://0.0.0.0:3001
      - MLFLOW_BACKEND_STORE_URI=postgresql://mlflow_user:mlflow_password@postgres:5432/mlflow
      - MLFLOW_S3_ENDPOINT_URL=http://minio:9000
      - AWS_ACCESS_KEY_ID=<ACCESS_KEY>              <-------------- Here
      - AWS_SECRET_ACCESS_KEY=<ACCESS_KEY_SECRET>   <-------------- And Here
      - MLFLOW_S3_IGNORE_TLS=true
    volumes:
      - ./mlruns:/mlruns
    depends_on:

[...]

Restart the docker containers:

docker compose down

docker compose up -d

Test

You can now test the configuration by running the notebook logging-first-model.ipynb