Bug 5614

Summary: TCK expects JobExecution to be updated as job runs
Product: jbatch Reporter: mminella
Component: TCKAssignee: ScottKurz
Status: RESOLVED FIXED    
Severity: major CC: issues
Priority: P5    
Version: 1   
Target Milestone: ---   
Hardware: All   
OS: All   
Whiteboard: tck_1.0

Description mminella 2013-12-03 23:08:18 UTC
This is essentially re-opening of 5371.  The TCK expects that a JobExecution is being updated behind the scenes instead of getting a new instance before checking the state.  The list below is what I have currently identified and will add tests to this list as I find more.

* JobOperatorTests#testInvokeJobWithUserStop()
* ParallelExecutionTests#testStopRunningPartitionedStep()
* ParallelExecutionTests#testStopRestartRunningPartitionedStep()
* StopOrFailOnExitStatusWithRestartTests#testInvokeJobWithUserStopAndRestart()

My guess is that most of the places JobOperatorBridge#*AndWaitForResult are called will need updates.
Comment 1 ScottKurz 2014-01-03 03:42:22 UTC
Fixed all these and will also mention that I found another similar issue with the testInvokeJobWithUserStopAndRestart() test.

If we're going to look at the BatchStatus before waiting for a terminating status, then we should allow we might see a status of STARTING (we'd been expecting STARTED).   Fixed this as well.