Functional Tests (Selenium) in Practice
summary
A practical Selenium setup in Java that runs 500 functional scenarios and reduces total execution time from 8 hours to 1 hour by parallelizing tests with TestNG and a remote Selenium server. Tests use a three-layer API (technical utilities, Page Objects, and Actions) plus a data-object repository that emits SQL to initialize the database; the talk also covers concrete solutions for AJAX waits, creating deterministic test data and mocking external web services, explains differences between Firefox and headless HTMLUnit, and addresses common flakiness and maintenance issues.
AI-written summary · about
chapters (10)
- 0:00Functional tests with Selenium
- 8:31Automated test maintenance
- 13:04Page Object architecture
- 21:54Test data scenarios
- 24:45Selenium app adjustments
- 26:52Parallel tests repeatability
- 31:34Browser selection limits
- 33:01Jenkins integration
- 36:32Test drawbacks and costs
- 39:29Summary
related talks
46mTest Automation with Selenium WebDriver, Java, and JUnit
37mSelenium In the Enterprise: What went right, what went wrong (so far)
34mUsing Selenium to Improve a Team's Development Cycle
36m3 ways not to use Selenium
52mStart writing good functional tests in Java with WebDriver (Mikalai Alimenkou, Ukraine), part 1
01h 14mContinuous Testing
49mStart writing good functional tests in Java with WebDriver (Mikalai Alimenkou, Ukraine), part 2
42mReliable and Maintainable Webdriver Tests