A large-scale effort to automate manual regression tests with an external team resulted in a slow, brittle, and unusable test suite.
The test suite was doomed by its three-hour runtime, constant failures, and complex local setup requirements, leading to the decision to delete most of them.
High internal quality, including robust testing, leads to faster feature delivery over time, making it a cost-effective strategy rather than an expense.
Assigning end-to-end responsibility for features and their corresponding tests to a single team creates accountability and ensures tests are maintained.
Successful end-to-end testing requires daily monitoring, debugging complex failure chains, and setting realistic goals for stability and execution time.
Modern architectures like microservices may benefit from alternative models like the test honeycomb or trophy, which place more emphasis on integration tests over unit tests.
Prioritize which tests to write by focusing on their return on investment, using heuristics like fixing production bugs and "embarrassment-driven" testing.
A solid testing practice reduces production bugs and deployment pain, which directly leads to less stress, less burnout, and higher job satisfaction.
The core principles for effective testing are to build them with features, ensure team ownership, monitor them closely, and focus on their return on investment.
56:49Jens Happe
28:12Luise Freese
26:20Augustin Gottlieb
43:07Golo Roden
47:22Christian Kühn
24:19Ramona Schwering
26:02Vincent Hennig & Anupam Krishnamurthy
23:40Ewald Verhoeven
.gif?w=240&auto=compress,format)
Jobs that call for the skills explored in this talk.
Think-cell
Foster City, United States of America
Saby New Compy
Karlsruhe, Germany
LORENZ Life Sciences Group
Vesterling Consulting GmbH
Dresden, Germany