Atomic and Non-Atomic Data-Driven Tests

When planning data-driven tests, you should first be aware of the following two different data-driven test types that are available in Silk Central:

Atomic data-driven tests: A single test result is generated for all data rows of your data source. This means that the test is only successful if the execution with every single data row is successful. If the execution with one data row fails, the whole test is marked as failed.
Non-atomic data-driven tests: Each data row of your data source is represented by a test of its own. This means that each data row produces a failed or passed test result. For example, if your data source is a spreadsheet with four rows, you will have the original test you created as a parent test in addition to four new child tests, one for each of the data rows.

For detailed information on creating data-driven tests, see Creating Data-Driven Tests.

The parent test created in this process does not have parameters associated with it, since it only represents a structuring instance for its child tests and no longer functions as an actual test. All values found in the data source will be listed in the parent test’s Data-driven Properties section.

When assigning a parent test to a requirement, links to requirements are only inherited when using atomic data-driven tests.

You can not assign the parent test of a non-atomic data-driven test to a setup or cleanup test execution, as such a parent node is treated as a folder. You can assign one of its child nodes though, and you can also assign an atomic data-driven test to a setup or cleanup test execution.

For detailed information on data source limits, see Data Source Limits.