PHT Station Backend

This project contains the implementation of the station API, workers for training models, as well as the configuration for the station airflow instance and related workers. A FastAPI REST API als well as a command line tool for train and station management can be found in the station directory.

Setup development environment

Checkout the repository and navigate to project directory.

git clone https://github.com/PHT-Medic/station-backend.git

cd station-backend

Prerequisites

Make sure the following ports required for the services are open or change the port mappings in the docker-compose.yml file.

• Postgres: 5432
• Redis: 6379
• Minio: 9000 & 9001 (Console)
• Airflow: 8080
• Blaze FHIR server: 9090
• API: 8000

Start services

Start the services for development using docker-compose.

docker compose up -d

Check the logs of the services to see if everything is running as expected.

docker compose logs -f

Configure station config for running the API in development mode

Copy the station_config.yml.tmpl file in the root directory to station_config.yml and adjust the values (especially configuring addresses and credentials for the central api)

# Configure authentication for central services
central:
  api_url: ""
  # Robot credentials for accessing central services, these can be obtained in the central UI
  robot_id: "central-robot-id"
  robot_secret: "central-robot-secret"
  private_key: "/path/to/private_key.pem"
  # optional password for private key
  private_key_passphrase: "admin"

######### some lines omitted #########

# Configures the address and credentials for the central container registry
registry:
  address:
  password:
  user:
  project:

Install python dependencies

Install dependencies using poetry. This will also create a virtual environment for the project.

poetry install --with dev

Run the station API

To run the station API with hot reloading, run the following command:

poetry run python station/app/run_station.py