DevOps中的测试层次与策略-优快云博客

摘 Azure DevOps Shift left to make testing fast and reliable

用例分类

Defining a test taxonomy is an important aspect to DevOps. Developers should understand the right types of tests to use in different scenarios, as well as what tests are required at different parts of the process. The categorization should also take into account the dependencies and time required for tests to run. Consider a system where tests are categorized across four levels:

L0 tests are a broad class of fast in-memory unit tests. An L0 test is a unit test to most people. That is a test that depends on code in the assembly under test and nothing else.
L1 tests might require the assembly plus SQL or the file system.
L2 tests are functional tests run against testable service deployments. It is a functional test category that requires a service deployment but may have key service dependencies stubbed out in some way.
L3 tests are a restricted class of integration tests that run against production. They require a full product deployment.

While it would be ideal for all tests to run at all times, it’s just not feasible at this time. Instead, teams can be more selective about where the line is drawn at which point in the DevOps process where each is run. While the expectation may be that developers always run through L2 before committing, a PR may automatically fail if the L3 test run fails, and the deployment may be blocked if L4 tests fail. The specific rules may vary from organization to organization, but enforcing the expectations for all teams within a given organization will move everyone toward the same quality vision goals.

ref: