Thanks David. Apologies for the lack of experience on this end, would the failure pipeline be able to be more of an "ignore" pipeline?
If the failure pipeline stopped future processor execution within the current pipeline and allowed the record to pass unmodified, that would definitely cover my goals.
I hadn't thought of what amounts to a "null failure pipeline", it's a bit like using exception processing paths for non-exceptional circumstances, generally considered an anti-pattern. But this isn't standard programming either. I just want to make sure I don't hand off something embarrassing or brittle to the next guy and establish good patterns as I learn.
Thanks!