Running with -n auto in PyTest - Build an Automation Script

Run pytest tests in parallel using -n auto

Preconditions (2)

Step 1: Open the terminal

Step 2: Navigate to the directory containing the test file

Step 3: Run the command: pytest -n auto

Step 4: Observe the test execution running in parallel

Step 5: Verify all tests pass successfully

✅ Expected Result: Tests run in parallel using all available CPU cores and all tests pass

Automation Requirements - pytest with pytest-xdist

Assertions Needed:

All test functions complete successfully

No test failures or errors occur

Best Practices:

Use pytest fixtures for setup and teardown

Avoid shared state between tests to prevent race conditions

Use -n auto to automatically use all CPU cores

Automated Solution

PyTest

import pytest

@pytest.fixture
def sample_data():
    return [1, 2, 3]

@pytest.mark.parametrize('num', [1, 2, 3])
def test_is_positive(num):
    assert num > 0

def test_sum(sample_data):
    assert sum(sample_data) == 6

# To run these tests in parallel, execute in terminal:
# pytest -n auto

This code defines simple pytest tests that can run independently.

The test_is_positive function checks numbers are positive.

The test_sum function checks the sum of a list.

Using pytest -n auto runs these tests in parallel using all CPU cores.

Fixtures and parameterization ensure tests are isolated and suitable for parallel execution.

Common Mistakes - 3 Pitfalls

{'mistake': 'Not installing pytest-xdist before using -n auto', 'why_bad': 'The -n option requires pytest-xdist plugin; without it, the command fails.', 'correct_approach': "Install pytest-xdist using 'pip install pytest-xdist' before running tests with -n."}

Sharing mutable global state between tests

{'mistake': 'Using print statements to debug without capturing output', 'why_bad': 'Parallel output can mix and become unreadable.', 'correct_approach': "Use pytest's capsys fixture or logging for better output management."}

Bonus Challenge

Now add data-driven testing with 3 different inputs and run them in parallel using -n auto

Show Hint