Neptune Deployment Tests

Tests worth running to confirm successful setup of a self-hosted neptune instance.

Tests are divided into:

1. basic_test.py that only checks minimal logging functionality & can be parametrized by number of metrics & steps logged
2. feature_tests.py with comprehensive checks for most logging & retrieval features (recommended for deployment validation)
3. performance_tests.py measuring logging throughput & retrieval performance (not implemented yet)

Some (<10%) of the tests will prompt you for maual validation actions, to be completed as a kind of post-deployment "checklist".

Quick start

Dependencies:

1. Create a project called deployment-tests in your workspace using neptune web app
2. Clone this repository & set up necessary environment variables:
   • export NEPTUNE_API_TOKEN=my-api-token, using your user token copied from your neptune web app
   • export NEPTUNE_PROJECT=my-workspace/deployment-tests, using workspace & project name from step 1

We recommend running the scripts using uv, which supports inline script metadata with package dependencies:

uv run basic_test.py
uv run feature_tests.py # this one will prompt you for 2-3 manual validation actions in the web app

You can re-try specifc tests using filtering:

uv run feature_tests.py -f fetcher # only runs tests that contain "fetcher" in their funciton name - no manual actions required

Alternatively, you can use pip & virtualenv:

python3 -m venv .env
source .env/bin/activate
pip install -r requirements.txt

And then run the tests using:

python basic_test.py
python feature_tests.py

To make test output less verbose, set NEPTUNE_LOGGER_LEVEL environment variable to warning.