Summary

This project intends to do 4 key goals:

1. Query MongoDB through Apache Superset 2.0.1
2. Create an embedded dashboard and a demonstration how to do it by a React App
3. Build full project by a simple and reusable docker-compose file so that quick start a Superset instance and Trino
4. Enable Celery worker for SQLLab and turn on the Global Async Query so that optimize performance for Dashboard loading

Steps

1. Clone current repository

   git clone https://github.com/zhaoyongjie/superset-containerization.git

2. Edit mongodb.connection-url in docker/mongodb.properties, fills up corresponding mongodb URI.
3. Startup Superset via docker-compose, notice that 1)There are some environment variables to declare in the docker/.env-non-dev. 2) the file superset_home/superset_config.py will override the config.py and read Database and Redis config from the docker/.env-non-dev.

   cd superset-containerization
   docker compose up -d
   

4. Use account admin/admin to login Superset.
5. Create a new Database in Superset, the SQLAlchemy URI is trino://admin@trino:8080/mongodb.
6. Add a Gamma role to admin user.
7. Generate Embedded dashboard uuid. Click embed dashboard on the three dots in a dashboard. Notice that Allowed Domains in the embed model could leave empty first.
8. Get guest_token for embedded dashboard. To make it easier to generate http requests, we use the httpie to request

   # request a JWT token for admin user
   echo '{"password": "admin", "username": "admin", "provider": "db"}' | http post http://localhost:8088/api/v1/security/login
   
   # request a guest token for embedded dashboard
   echo '{"user": {"username": "admin"}, "resources": [{"type": "dashboard", "id": "<fill up a dashboard id>"}], "rls": []}' | http post http://localhost:8088/api/v1/security/guest_token/ -A bearer -a "<fill up JWT token from previous step>"

9. fill embedded uuid and guest token in embedded_superset/src/App.jsx
10. run npm i in embedded_superset
11. run npm run build in embedded_superset
12. open a http server in embedded_superset/dist

    cd embedded_superset/dist
    python3 -m http.server 5000

13. open a browser and explore localhost:5000, the embedded dashboard should be showed.