Test Case Traceability

[Beth]

I read through this thread (http://www.sparxsystems.com/cgi-bin/yabb/YaBB.cgi?num=1254125691/0) with interest as it seems to be related to what I'm trying to do - provide some traceability between my Requirements and Use Cases to Test Cases.  It is not obvious to me on how to do this with EA.

What we have are our System Requirements, which we are modeling in the tool as the Requirements.  We decompose those further into Use Cases.  I can provide bi-directional traceability between the Reqs and the Use Cases without any problem.

Now I want to create TestCases within EA that will allow me to see the traceability to either the Reqs or Use Cases.

We have "System" Test Cases which make up our System Verification Test, which need to test at the Requirement level.

We have "Integration" Test Cases which test the various scenarios of a Use Case.

I see how I can create and document the different types of Test Cases with the Testing View (that has the 5 tabs).  But I don't see how I can map between these Test Cases and either the Reqs or Use Cases.

When I look at the diagram in the post referenced above (the one that models the scenarios as operations on the Use Case) - am I correct in assuming you are doing this entirely with the Use Case modeling, and you are NOT defining Test Cases (as defined by the Test View)?

I've read through the Testing Support page on the web, and also read the help on Testing.  But I don't see anything that tells me how to setup the models to provide that traceability.

Can anyone give me some pointers?
Thanks,
Beth

PS. We are currently using v7.1.833

[son-of-sargasso]

Hi Beth,
Not ignoring you, still thinking.

am I correct in assuming you are doing this entirely with the Use Case modeling, and you are NOT defining Test Cases (as defined by the Test View)?

Yes, I have found that the testing view is not functionally rich enough to provide what I need in test management.  Some missing principles are retaining a test result history, links to defect management and a meaningful way to subset testcases by release.

Regarding traceability, for functional testing (no matter what test level) I have found that relating test cases to use case is the most meaningful way for most test observers.  Auditors (who don't understand what exactly UML footballs mean) can be told gently (using an axe if necessary) that if the test cases provide adequate coverage of the usecases then they must cover the requirements.

bruce