We have a parameterized job to run database integration tests. Each 
distinct parameter corresponds to a database brand/version/... The regular 
build ends with triggering jobs for each parameter value.

The problem is that the parameterized job aggregates the results of all the 
different database types. And will consider the job successful when the 
last job in the series succeeds. Regardless whether other instances in the 
series were not. Which doesn't make much sense - in my case that is.

I've tried creating jobs whose sole purpose is to trigger one instance of 
the parameterized job. Hoping it would kind of aggregate the results of 
that one specific parameterized job. But it didn't. Appears triggering a 
job is an asynchronous affair; the delegating jobs will always succeed.

What I'm looking for is to run *exactly* the same tests for each database 
we support. And keep the results grouped per database type. Without 
duplicating the job definition. Otherwise we're risking tests to be 
different because of mistakes.

Is that currently possible with Jenkins ?



Reply via email to