All data is stored on the chairs MariaDB database as specified in `Settings -> CI/CD -> Variables` and is accessible from the Petra Server only. However, the PhpMyAdmin allows access via browser. For the login, use same credentials as the backend does.

## DB Schematics

The database has the following entities and relationships with a cycle centric approach. The diagram can be forked at [dbdiagram.io](https://dbdiagram.io/d/603ffba5fcdcb6230b227581)

![Petra_DB_Schema](../uploads/f9c067a69b016cef7aaeb097892b09ad/Petra_DB_Schema.png)

## Create Database

If no tables are present in the database e.g. after changing the database, flask automatically creates entities and relationships. For this to work, connect to a running docker backend container and execute
```bash
$ flask db upgrade
```
to create the tables and relations via SQLAlchemy and
```bash
$ flask shell
>>> from app import routes
>>> routes.setup_endpoint()
```
to add the default values e.g. Admin team.

## Upgrade Database

The database entities and relationships are defined at `backend/app/app/models.py`.
However, after changing the models the database requires an upgrade handled by [flask migrate](https://github.com/miguelgrinberg/flask-migrate).
Login to a running docker backend container that has access to the database (should already be the case, though) and execute the following to upgrade the database:
```bash
$ flask db migrate -m "<OPTIONAL MESSAGE>"
$ flask db upgrade
```
This, on the one hand, identifies changes applied to the `models.py` with an optional message and, on the other hand, persists these changes to the DB.