I created some automated tests using Selenium IDE first, then I converted them into JUnit 4 WebDriver format. After making the necessary changes to get them working properly as JUnit tests, I now want to run these tests in my continuous integration environment.
Since CI servers don’t have a display, I need to configure my JUnit tests to use a headless browser. I’ve seen some general discussions about headless testing, but I’m specifically looking for guidance on how to set this up with JUnit WebDriver tests.
What would be the recommended way to handle this? I’m open to any suggestions or best practices for running JUnit Selenium tests without a GUI in a CI environment.
firefox headless is solid if u dont want chrome. just add --headless to firefoxOptions and ur set. ive run it on jenkins for months - zero issues. way more stable than phantomjs, which is dead anyway.
I ran into the same issues moving from IDE to CI. What worked for me was building a WebDriver factory that auto-detects the environment. Just check system properties or environment variables to see if you’re in CI, then spin up the right driver config. For Chrome, I had to add --disable-gpu and --window-size with --headless to fix some weird rendering bugs. Heads up though - headless can act differently than regular browsers, especially with JS-heavy pages. Test your converted IDE tests thoroughly in headless before pushing to CI. But the speed boost is worth it - my suite runs 40% faster headless.
To run JUnit WebDriver tests in headless mode, configuring ChromeOptions is crucial. I recommend using flags like --headless, --no-sandbox, and --disable-dev-shm-usage, as they significantly enhance performance in CI environments. It’s also beneficial to implement a system property check to easily switch between headless and regular modes. Be cautious with screenshot functionality in headless mode, as it can disrupt how images are captured or displayed. Overall, this setup has sped up my tests by approximately 30%.