Skip to content

docker-deployment.md

Latest commit

 

History

History
64 lines (43 loc) · 1.54 KB

docker-deployment.md

File metadata and controls

64 lines (43 loc) · 1.54 KB

Docker Deployment

Building the Docker Image

Build the image:

docker build -t assets-api . -f Dockerfile

Verify the build:

docker images

Running the Container

Run the Docker container, making sure your AWS credentials are configured in the .env file:

docker run -p 8000:8000 -v $(pwd)/.env:/app/.env.api:ro assets-api

Testing the API

Run a classification job on a PDF file. Make sure the file exists in the configured S3 bucket and folder before starting.

# Run classification
curl -X POST http://127.0.0.1:8000/v1/ \
  -H "Content-Type: application/json" \
  -d '{"file": "YOUR_OWN_FILENAME.pdf"}' -i

# Collect results
curl -X POST http://127.0.0.1:8000/v1/collect \
  -H "Content-Type: application/json" \
  -d '{"file": "YOUR_OWN_FILENAME.pdf"}' -i

Docker Compose

The repository includes a docker-compose.yml for convenience:

docker compose up

MinIO

If you are using a local MinIO instance for file serving, localhost inside the container refers to the container itself, not your host machine. Update the endpoint in .env so the container can reach MinIO running on your host:

LOCAL_S3_ENDPOINT="http://host.docker.internal:9000"

Run from Github Packages

The built Docker image for development is present on the repository under the Pakages. To run it locally use:

docker pull ghcr.io/swisstopo/swissgeol-assets-dataextraction-api:edge
docker run -p 8000:8000 -v $(pwd)/.env:/app/.env.api:ro ghcr.io/swisstopo/swissgeol-assets-dataextraction-api:edge