How can I run all the JUnit tests of a spring boot application two times?
First time with a custom parameter (say a.b="one") in application.yml and the second time with a.b="two"
What needs to be configured?
I assume changes will be necessary in JUnit test classes and in pom.xml
I can mark a
#WebMvcTest
With
#TestPropertySource(properties = "a.b=one")
or with
#TestPropertySource(properties = "a.b=two")
and it will use that option during execution. But how to run it twice with these two different options? Is it somehow doable with
org.junit.jupiter.params.ParameterizedTest ?
Related
I have an integration test that's using #SpringBootTest to spin up a Spring application context that's testing a simple Spring Boot app. I'm using Spock for writing the tests, my build tool is Maven.
I'm looking for a way to run the same test class multiple times with different configurations for the test (I have a set of config options and I need to ensure consistent behavior for a certain permutation of config options). The first idea I had was to use profiles to define the exact permutation, maybe it could also work by using #TestPropertySource in some way. However, I don't see any way to run a test class multiple times, using different configurations each time.
I know that I could run all tests with a given profile, but in my case I want to only apply the different configs to certain test classes.
I can also use a where block to repeat spock tests as described here, but that doesn't allow me to switch spring configurations for each run
The easiest way is to use simple subclasses, i.e., you define all your tests in an abstract base class and then subclass it for every variation and add the necessary annotations to the subclasses. This approach works well if you only have a limited set of variations, and provides good reporting feedback as every variation is reported as its own specification.
Problem: When an integration test fails, it often not delete his test data (I mostly use DbUnit for that). Because of this, other tests will also fail (for example because they want to delete their test data, but it is referenced by data which shouldn't exist at this point --> test will fail).
Is there a possibility that Jenkins use for every single test or for every Java test class a fresh/clean database with the init data (I don't need a complete empty database), that my integration tests are more independent?
Use a spring.jpa.hibernate.ddl-auto=create-drop property or annotate your test class with #Transactional.
Say I have 10 spring boot test class (annotated with #RunWith(SpringRunner.class) and #SpringBootTest)
Each test needs to launch spring container for like 10 seconds, although the container might do the same init.
So I may need 100 seconds for "mvn test".
Is there a way I can group my 10 test class into 1 suite, and let the container only start once.
So I can:
Only run the suite for "mvn test". (with proper naming for individual test class)
Optionally run individual test in IDE.
Spring uses Cache Management to cache the Application Context between tests:
By default, once loaded, the configured ApplicationContext is reused for each test. Thus, the setup cost is incurred only once per test suite, and subsequent test execution is much faster. In this context, the term “test suite” means all tests run in the same JVM — for example, all tests run from an Ant, Maven, or Gradle build for a given project or module. In the unlikely case that a test corrupts the application context and requires reloading (for example, by modifying a bean definition or the state of an application object) the TestContext framework can be configured to reload the configuration and rebuild the application context before executing the next test. (https://docs.spring.io/spring-framework/docs/current/spring-framework-reference/testing.html#integration-testing)
So this mechanism tries to execute your integration tests on an already running Application Context if possible. As you see multiple Application Context launches, this indicates your tests somehow use a different setup e.g. different profiles active, test properties, MockBeans etc.
The Spring documentation provides an overview on which indicators it puts an Application Context in its cache: https://docs.spring.io/spring-framework/docs/current/spring-framework-reference/testing.html#testcontext-ctx-management-caching
If you e.g. don't change any test property for your integration tests, Spring can run all of them on only one Application Context and be extremely efficient.
Another indicator for your current behaviour might be the use of #DirtiesContext which leads to a fresh Application Context after your test executes.
I have a #Service class where i'm caching some table data. I don't want those queries to run while building mvn install. Is there a way to ignore the file while building and it only execute when i start the server ?
It's a spring-boot application.
Here is background of my issue. I have initialized the spring boot app from http://start.spring.io/ site, which actually adds dummy application test file with SpringBootTest annotation and default contextLoads() with Test annotation, with an intention to initialize and execute all test cases, which needs to initialize and execute all code base. In my opinion this is not required, as we can have respective Test classes per controller/manager, which will give more controlled environment to hook up your Test setups and executions.
I have removed the default application Test file and included respective test classes for code coverage and quality. This way my beans are not executed at server startup time rather build time.
I am now using Spring and JUnit, and I use #DirtiesContext to get my test cases independent of each other, but I get OutOfMemoryError each time I run my test case.
Is there any that I can config the memory use of #DritiesContext ?
Thank you!