Closed vezzi closed 9 years ago
I'll look into this.
Well, now I've done some thinking about this, and I think that this should be the expected behavior. If you send a exit signal to the application and it finished gracefully (as Piper will normally do) it should give you a 0 exit flag, because there was really no error. It should only return a non-zero exit flag if there was an error - that is the pipeline it self failed or the jobs in it failed. Does that sound reasonable to you?
Yes it is. I have already worked around this "non problem" checking for the existence of the results. If no result is present then we will set the run to fail
:thumbsup:
I am not entirely sure but I noticed this behaviour in piper: I run piper through the ngi_pipeline, if the ngi_pipeline fails on its own then the piper instance continue to run probably because no ctr+c signal is sent (and this is fine).
The problem is that if I ctr+c the ngi_pipeline then piper captures the signal and exist. However, in this case the return code that he gives is "0"
Is this possible? Should I explore this more or you have some explanation for this behaviour?
The return code is important pipeline side to know if we can consider a run fail or not