GeoAPI is a restful API to create geospatial features in a PostGIS database. Users create a map "project" then can add features to it. The development docker-compose file has 3 containers:
- a PostGIS database exposing 5432,
- the api which exposes port 8000 behind gunicorn
- a nginx server to serve static files and proxy to the api, running on port 8080.
See https://github.com/TACC-Cloud/hazmapper which is an associated viewer application.
Environment variables are used to define the tenant's service accounts which are needed to access metadata and some user information. An .env file for developers can be found on UT Stache.
The API is built with flask and flask-restplus. It is running in its own container under gunicorn on port 8000
make build
make start
docker exec -it geoapi python initdb.py
You need a Tapis token for the appropriate tenant.
export JWT=your_access_token_stringTo create a new "map" project, send a POST request:
curl -X POST -H "Content-Type: application/json" -H "X-Tapis-Token: $JWT" http://localhost:8000/projects/ -d '{"name": "Test Project", "description": "This is a test project."}'
To view all projects, including the newly created one, send a GET request:
curl -v -H "Content-Type: application/json" -H "X-Tapis-Token: $JWT" http://localhost:8000/projects/
See https://github.com/TACC-Cloud/hazmapper for details.
These are useful steps to follow when there are changes to the database model.
First, apply migrations:
docker exec -it geoapi alembic upgrade head
Note: The above step is also automatically performed when running initdb.py
Then, create migrations:
docker exec -it geoapi /bin/bash
# determine a description for the migration like 'add_user_email_column'
alembic revision --autogenerate -m "add_user_email_column"
# Then:
# - remove drop table commands for postgis
# - add/commit migrations
Run directly in your running containers:
# then run tests in api
docker exec -it geoapi bash
APP_ENV=testing pytest
# then run tests in worker
docker exec -it geoapiworkers bash
APP_ENV=testing pytest -m "worker"
Information on Kubernetes configuration for production and staging environments can be found in the kube/README.md including information on kube commands and Jenkins deployment workflows.
The python package can be found at PyPi. More details can be found in Python Client