[openstack-dev] [telemetry] Moving Gnocchi out

David Moreau Simard dms at redhat.com
Wed Mar 22 16:34:32 UTC 2017

I'm really curious about this decision too.
Not so much about why the project wants to move out but more about
what it plans on doing in terms of contribution (code, issue)

I happen to drive a project that's not OpenStack-specific either: ARA [1].
ARA was first created in GitHub and was "incubated" there until we
felt it was "good enough" to be proposed as an OpenStack ecosystem

We chose the "OpenStack workflow" for two main reasons:
- The original authors were already intimate with it and we were very
satisfied with the rigid process it provided
- ARA would be used by different OpenStack projects and it would be a
good fit to be a part of the "family"

I did not find it hard to find *users* outside of the OpenStack
bubble, however, I still felt I needed to document a "FAQ" [2] about
how, yes, the project can be used outside of OpenStack.
It is definitely challenging to find contributors outside the
OpenStsack ecosystem, however. Even with an attempt at providing
simplified contribution guidelines [3].

Signing up for Launchpad and OpenStackid accounts, learning
git-review, setting up and using Gerrit and tracking things in
launchpad or storyboard are things we take for granted.
Whatever way we put it, though, it's a higher barrier to entry than
just browsing the GitHub repository and filing issues or creating pull
requests there.

So, what's the alternative ? Use the GitHub workflow ?
How well is this working out for projects that attracts (or intends to
attract) a lot of users and developers ?

Look at the Ansible GitHub repository [4] for an extreme case: 2600
contributors, more than 1700 issues and almost 1000 pull requests.
How do you make sense out of that ?

Ansible has had to create a bunch of custom software to wrap around
the workflow.
Triaging bots [5], custom tools [6] to sift through the amount of
content they have and so on.

I'm not saying the OpenStack workflow is better than the GitHub one --
just that there are pros and cons that the project must weigh based on
it's priorities and resources.

That said, I'll re-iterate that I'm really curious on what Gnocchi
intends on doing.

[1]: https://github.com/openstack/ara
[2]: http://ara.readthedocs.io/en/latest/faq.html#can-ara-be-used-outside-the-context-of-openstack-or-continuous-integration
[3]: http://ara.readthedocs.io/en/latest/contributing.html
[4]: https://github.com/ansible/ansible
[5]: https://github.com/ansible/ansibullbot
[6]: http://jctanner.mynetgear.com:5000/issuesearch/programmer-defection-vacuum/created_at/desc

David Moreau Simard
Senior Software Engineer | Openstack RDO

dmsimard = [irc, github, twitter]

On Mon, Mar 20, 2017 at 12:57 PM, Ian Cordasco <sigmavirus24 at gmail.com> wrote:
> -----Original Message-----
> From: Chris Friesen <chris.friesen at windriver.com>
> Reply: OpenStack Development Mailing List (not for usage questions)
> <openstack-dev at lists.openstack.org>
> Date: March 20, 2017 at 11:39:38
> To: openstack-dev at lists.openstack.org <openstack-dev at lists.openstack.org>
> Subject:  Re: [openstack-dev] [telemetry] Moving Gnocchi out
>> On 03/20/2017 10:10 AM, Chris Dent wrote:
>> > On Mon, 20 Mar 2017, Thomas Goirand wrote:
>> >
>> >> I really don't understand why the Telemetry team insists in being
>> >> release-independent, out of big tent and such, when the reality is that
>> >> all of released Telemetry components are *very tightly* bound to a
>> >> specific versions of OpenStack. IMO, it doesn't make sense upstream, or
>> >> downstream of Telemetry.
>> >
>> > This simply isn't the case with gnocchi. Gnocchi is an independent
>> > timeseries, metrics and resources data service that _happens_ to
>> > work with OpenStack.
>> >
>> > By making it independent of OpenStack, its ability to draw
>> > contribution and engagement from people outside the OpenStack
>> > community increases. As a result it can become a better tool for
>> > more people, including OpenStack people. Not all, or even many, of
>> > the OpenStack projects are like that, but gnocchi is. More eyes,
>> > less bugs, right?
>> I'm curious why being independent of OpenStack would make it more attractive.
>> Is the perception that requiring people to sign the Contributor Agreement is
>> holding back external contribution? Or is it just that the mere idea of it
>> being an OpenStack project is discouraging people from getting involved?
>> Just as an example, if I want to get involved with libvirt because I have an
>> itch to scratch the fact that it's basically a RedHat project isn't going to
>> turn me off...
> Contributing to OpenStack is intimidating, if not utterly
> discouraging, to people unfamiliar with CLAs and Gerrit. There's a lot
> of process that goes into contributing. Moving this to a friendlier
> (if not inferior) developer platform makes sense if there is interest
> from companies not interested in participating in the OpenStack
> community.
> --
> Ian Cordasco
> __________________________________________________________________________
> OpenStack Development Mailing List (not for usage questions)
> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

More information about the OpenStack-dev mailing list