Choosing a Database Engine

Choosing a Database Engine

All three common options — SQLite, PostgreSQL, MySQL — are free. The database software itself is free in every case; cost only appears with cloud managed databases (e.g. AWS RDS), which is hosting cost, not a license fee. Postgres also has free hosting tiers (Supabase, Neon), so deploying can stay at $0.

	SQLite	PostgreSQL	MySQL
License	Public domain	Open source (permissive)	GPL (Community); MariaDB is a fully-free fork
Setup	None (single file)	Server install	Server install
Concurrent writes	Weak (one at a time)	Strong	Strong
Django fit	Default	Best (de facto standard)	Fine
Deployment	Poor*	Excellent	Good
Best for	Local dev / learning	Production	(Postgres covers it)

*On many hosting platforms the filesystem resets on redeploy, which can wipe the SQLite file entirely — so it's unsuitable for production.

Recommendation for this project

Start on SQLite, switch to PostgreSQL for deployment.

• During development → SQLite. Zero setup; Django uses it by default.
• For deployment → PostgreSQL. Best Django fit, and it handles the daily batch and concurrent users without issue.
• MySQL → not bad, but in the Django ecosystem Postgres is the standard, so there's little reason to pick it. Postgres also aligns better with the ORM (e.g. JSONField, constraints).

Django makes swapping engines easy — only the DATABASES block in settings.py changes:

# Development (SQLite)
DATABASES = {
    "default": {
        "ENGINE": "django.db.backends.sqlite3",
        "NAME": BASE_DIR / "db.sqlite3",
    }
}

# Production (PostgreSQL) — requires: pip install psycopg2-binary
DATABASES = {
    "default": {
        "ENGINE": "django.db.backends.postgresql",
        "NAME": os.environ["DB_NAME"],
        "USER": os.environ["DB_USER"],
        "PASSWORD": os.environ["DB_PASSWORD"],
        "HOST": os.environ["DB_HOST"],
        "PORT": "5432",
    }
}

Caveat: developing on SQLite but deploying on Postgres can surface subtle differences (e.g. case sensitivity). If that matters, run Postgres locally too (usually via Docker) to match dev and production. For learning, SQLite is the lighter way to start.