08-11-2016 11:32 AM - edited 08-11-2016 11:36 AM
When running a sequence in a new execution (to obtain a seperate report for each test sequence) the status indicates pass even though a run-time error has occurred. The system is set to ignore errors. When running the same sequence in isolation or not in a new execution the status indicates failed. Is there a way to avoid this without using workarounds http://digital.ni.com/public.nsf/allkb/1CEA4B6BECD26D53862575FC007ED8E9 inserted into every sequence and sub-sequence?
08-11-2016 11:42 AM
Trying to follow your logic here: Basically you are spawning a new execution from MainSequence? The new execution encounters an error. You expect MainSequence to report a Failed?
By spawning a new execution you now have a different SequenceContext than what is being put on the report. So the key here is that eventhough your new execution errors or "fails", MainSequence context does not see that. You have to set tha MainSequenc context to be "failed". One way to do this is to pass the sequencecontext into you new execution and then set it's status to Failed.
Regards,
08-12-2016 02:22 AM
The point with New Thread and New Execution is that the newly spawned thread/execution runs in parallel to the calling thread. So the sequence call does not wait (by default) if the new thread/execution fails. You will only get a Done for the sequence call step as the call has been done....
Norbert
08-12-2016 05:55 AM
08-26-2016 05:42 AM
Hi Jiggawax,
So sorry for the late reply, i had some unexpected time off.
Thanks for correctly identifying the issue.
We used the SequenceFilePostStepRuntimeErrror call-back in the end.
This then ensures the top-level report indicates fail correctly.
Rgds, Torpedotown
08-26-2016 05:46 AM
Hi Norbert_B,
So sorry for the late reply.
I take your point regarding the parallel execution.
Many thanks,
Rgds, Torpedotown