A Processing Group would serve as a reusable unit, which then has version
control applied to it. You would model your flows to abstract interactions
with e.g. in/output port objects.
They also mentioned that nested PG can be versioned separately, looks like
git submodules behavior, but I haven't tried multi-level versioned PGs yet.
On Mon, Feb 19, 2018, 4:46 PM Georg Heiler <georg.kf.hei...@gmail.com>
> How can I modularize and reuse parts of my nifi flows?
> I.e. Have a shared logging and error handling strategy which can be
> centrally configureed for all processor groups.
> Best Georg