What is it?
In Software Testing, integration tests are made when Unit Testing is done and units are ready to be integrated into the system. Integration testing only happens when all units are have passed through Unit Testing.
The objective is to check any errors or faults between different units when they interact among themselves in the integration.
Methods for integration tests
-
Big-bang integration
The simplest approach: all modules are put together and then tested at once. Typically used for smaller projects with lower degrees of interdependence.
-
Bottom-up integration
Modules at lower lever are tested with higher modules until everything is tested, e.g. interface tested with driver or databases.
-
Top-down integration
Simulate the behavior of lower level modules that are not yet integrated, while testing high level modules, and then as they are integrated, goes down a level.
-
Mixed integration
Follows both bottom-up and top-down methods. Lower and higher modules are tested in parallel, until they meet in the middle.