How Your Tests Are Run

Endform provides a distributed test runner optimized specifically for Playwright, designed to run your end-to-end tests faster and more efficiently. Here’s a breif explanation of how our test runner works.

How It Works

Step 1: Dependency Analysis

• When you start your tests using our CLI, Endform first analyzes your project’s js/ts test files.
• It calculates the exact JavaScript files required for each test to run.
• The CLI determines common dependencies across your entire test suite.
• The set of tests to run, and the dependencies required for each test, are then sent our test runner service.

Step 2: Distributed Execution

• Endform starts a fully isolated machine for each test. Your tests run in dedicated Firecracker MicroVMs.
• Each machine receives only the specific test files and dependencies it requires.
• Each machine only ever runs one attempt of a test, the test runner can immidately schedule more than one retry of a test when the first run fails.

Step 3: Result Coordination

• Each test collects a playwright blob-report containing the result of the test.
• We always send your blob reports back to the cli. Once all your tests have finished, we run playwright’s merge-reports command for you. This means that all of your existing playwright reporters will work as expected.
• We also collect the results of your tests and the traces they produce for you to view in the dashboard.

Benefits

• Fast and Scalable: Run hundreds of tests in parallel without being limited by browser or network bottlenecks.
• Efficient: Minimized resource usage by isolating each test and its dependencies.
• Local Convenience: Execute large test suites effortlessly from your local machine.