Connect an application to a database safely

Database safety starts before connection pooling. Query values must remain data, and every tenant-scoped read must include trusted tenant context.

Execute a parameterized query

The integration lab uses an in-memory SQLite database.

from integration_cases import safe_database_case

result = safe_database_case()
assert result["malicious_rows"] == []
assert result["valid_rows"] == [("U1", "A")]
assert result["parameterized"] is True
assert result["tenant_filter"] is True

The injected ID is treated as a value, and the valid lookup is tenant-scoped. This does not prove production database security.

Use a least-privilege service account, encrypted connection, secret reference, bounded pool and timeouts. Parameterize values; allowlist dynamic identifiers. Wrap related writes in transactions and map uniqueness/conflict errors deliberately.

Manage schema through reviewed migrations, backup before risky changes and test upgrade/rollback or forward recovery. Avoid logging full SQL with literals or connection URLs. Monitor pool exhaustion and slow queries with safe summaries.

The FDE for Freshers course connects SQL modelling to safe application adapters.

Exercise

Add tenant-scoped create/read operations to a local database. Test injection strings, cross-tenant IDs, transaction rollback and uniqueness conflict.

Continue learning

This article is part of the FDE integration and deployment foundations sequence. Use the neighbouring tasks when you need the prerequisite or the next application.

• Review the prerequisite or neighbouring task in Build a FastAPI endpoint with validated inputs.
• Continue with Design a webhook receiver that tolerates duplicate events.

Reference: Python `sqlite3` parameter substitution.