In this conference talk, we will explore a cutting-edge approach to testing in a microservice architecture. With over 300 synchronous and asynchronous API endpoints running in a shared environment with cross-dependencies, traditional testing methods have become increasingly challenging. To address this, we have leveraged newer technologies to split our API testing into five levels, increasing our overall confidence. These levels include mocked black box testing, temporary namespaced API in our CI environment, post-deployment tests, environment stability tests, and data explorer tests. Each level focuses on specific functionality and acceptance criteria, allowing us to thoroughly test the APIs. We will demonstrate how we have orchestrated these tests using Kubernetes and CI configuration, and how we have employed a custom data extractor to find strange data sets to use as test data for data explorer tests. Join us to learn how this approach has changed the way we test and look at APIs, and to discuss the pros and cons of using this method.
Key Takeaways:- API automation at different layers of the stack
- Using development technology in testing
- Mocking frameworks to proxy requests