Terrific! Thanks for fixing this, Joe. This will help us catch DRT breaks in the controls tests more easily.
Eddie
Joe Pemberton (JIRA) wrote:
[ http://issues.apache.org/jira/browse/BEEHIVE-448?page=all ]
Joe Pemberton resolved BEEHIVE-448:
-----------------------------------
Resolution: Fixed
Should be fixed with revision 168006.
controls TCH test infrastructure doesn't fail the build when the tests fail ---------------------------------------------------------------------------
Key: BEEHIVE-448 URL: http://issues.apache.org/jira/browse/BEEHIVE-448 Project: Beehive Type: Bug Components: Controls Versions: V1Beta Reporter: Eddie O'Neil Assignee: Joe Pemberton Priority: Critical Fix For: V1
When the controls DRTs run and tests are aborted, the test run does not fail. This produces a false positive success on the run. An example can be seen here: http://beehive01.bea.com/downloads/20050316052618/general/build.out Scroll down to look at the output of TCH and see three aborts. These aborts stack traced here: http://beehive01.bea.com/downloads/20050316052618/controls/tch.cout Probably a good idea to fail the build on DRT runs that don't pass so that failures are obvious.
