Running Tests¶

Fast checks¶

Run these checks before changing integration code:

uv run ruff check .
uv run pytest -m "not integration and not slow"

The suite includes harness unit tests, distribution checks, schema registry checks, artifact tests, parser tests, offline checks, and non-container durability baselines.

All non-container tests¶

uv run pytest tests suites -m "not integration"

The performance benchmark test is small by default. The 500 migration and scale tests require their explicit environment variables.

Integration tests¶

DBWARDEN_HARNESS_RUN_INTEGRATION=1 uv run pytest -m integration

Run one suite while developing:

DBWARDEN_HARNESS_RUN_INTEGRATION=1 uv run pytest -q suites/round_trip/test_generate_models_integration.py

Use -k to select a backend. The backend matrix uses this same mechanism to avoid starting unrelated containers in each job:

DBWARDEN_HARNESS_RUN_INTEGRATION=1 uv run pytest -m integration suites/round_trip -k clickhouse

Select provider versions¶

The provider lifecycle suite supports DBWARDEN_HARNESS_BACKEND and DBWARDEN_HARNESS_VERSIONS:

DBWARDEN_HARNESS_RUN_INTEGRATION=1 \
DBWARDEN_HARNESS_BACKEND=postgres \
DBWARDEN_HARNESS_VERSIONS=14,17 \
uv run pytest -m integration suites/round_trip/test_providers.py

Slow suites¶

DBWARDEN_HARNESS_RUN_SCALE=1 uv run pytest -m slow suites/performance

The 500 migration benchmark is opt in:

DBWARDEN_HARNESS_RUN_500_MIGRATION=1 uv run pytest suites/performance/test_convergence_500.py -s

Artifacts¶

DBWARDEN_HARNESS_ARTIFACT_DIR=artifacts \
DBWARDEN_HARNESS_RUN_INTEGRATION=1 \
uv run pytest -m integration suites/round_trip

On failure, the artifact directory contains command output, temporary project files, generated migrations, model state, provenance, and provider logs when the provider is available to the pytest hook.