nifi-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jeff Zemerick <jzemer...@apache.org>
Subject Re: [DISCUSS] Bringing MiNiFi into NiFi
Date Thu, 21 May 2020 19:22:32 GMT
Apologies if this has been discussed previously. I've not kept up on the
Mi/NiFi progress as much lately.

When I think of two projects being combined I usually consider the
connection between the lifecycles of the two projects. One thing I always
liked about MiNiFi being separate was its ability to evolve outside the
NiFi lifecycle. New features, enhancements, and fixes could be released as
needed. It also had its own voting process for releases. But it certainly
had its drawbacks, too.

With the two projects being combined, is there any fear of negative effects
to MiNiFi's development being tied to NiFi's release cadence? Will it be
possible to do a MiNiFi release outside of a NiFi release? Or any desire to?

I'm not at all proposing not to merge the projects because there's a lot to
gain as stated in the initial email, but I would like to see MiNiFi be able
to maintain some of that independence and flexibility, if possible.

Thanks,
Jeff

On Thu, May 21, 2020 at 1:54 PM Joe Witt <joe.witt@gmail.com> wrote:

> Yes please!
>
> On Thu, May 21, 2020 at 1:53 PM Andy LoPresto <alopresto@apache.org>
> wrote:
>
> > Matt,
> >
> > I think this is a great idea, and I think it could also provide a bridge
> > to the reduced kernel build of NiFi with separate extensions that the
> > extensions registry will ultimately offer. Once the MiNiFi and NiFi
> > assemblies are complete, it should be possible to add a third which does
> > include the UI/API, but does not include the majority of specialized
> > processor NARs, etc.
> >
> > Andy LoPresto
> > alopresto@apache.org
> > alopresto.apache@gmail.com
> > He/Him
> > PGP Fingerprint: 70EC B3E5 98A6 5A3F D3C4  BACE 3C6E F65B 2F7D EF69
> >
> > > On May 21, 2020, at 10:49 AM, Pierre Villard <
> > pierre.villard.fr@gmail.com> wrote:
> > >
> > > Nothing useful to add other than I know this is going to be a lot of
> > work,
> > > but this is GREAT!
> > >
> > > Thanks,
> > > Pierre
> > >
> > > Le jeu. 21 mai 2020 à 19:44, Matt Burgess <mattyb149@apache.org> a
> > écrit :
> > >
> > >> All,
> > >>
> > >> I'm currently working on MINIFI-422, which aims to bring the MiNiFi
> > >> code into the NiFi codebase and have the MiNiFi product be a
> > >> specialized assembly of NiFi. Picture two different Maven profiles
> > >> that create a NiFi assembly or a MiNiFi assembly, each including the
> > >> common elements but excluding those things that aren't needed, such as
> > >> MiNiFi being "headless" and not including Jetty or the UI.
> > >>
> > >> This will be a lot of effort and very invasive to NiFi itself, so I
> > >> created a MINIFI-422 branch (based on the current master) and pushed
> > >> that to the Apache NiFi Github repo. I figure that way we can carve
> > >> out individual tasks and write PRs against that branch, rather than
> > >> duplicating code in the meantime and having MiNiFi show up little by
> > >> little in the NiFi codebase.
> > >>
> > >> My first task to that end is to continue Aldrin's work on MINIFI-488.
> > >> I originally had a PR against master for this subtask, but after
> > >> discussing with Aldrin we thought it would be better to have a
> > >> separate branch and incrementally add MiNiFi functionality until we're
> > >> ready for a big PR to bring it all into NiFi.
> > >>
> > >> We still want to keep up with NiFi master, so the branch would be
> > >> subject to force-pushes, rebases, etc. For that reason I'd like to ask
> > >> that anyone working on that branch please reach out to me
> > >> (mattyb149@apache.org) so we can coordinate and collaborate, to
> ensure
> > >> we're not stepping on each other's toes :)
> > >>
> > >> Also happy to discuss alternate approaches or any other questions or
> > >> concerns you may have.
> > >>
> > >> Regards,
> > >> Matt
> > >>
> >
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message