[openstack-dev] [Neutron] Evolving the stadium concept

Gary Kotton gkotton at vmware.com
Sun Feb 7 02:31:46 UTC 2016


At times I feel like it is the 80’s and we are witnessing the Heysel stadium disaster. My understanding from the whole thread and the patches in review for the stadium criteria is that accidents happen and the community is trying to prevent that. Honestly I am really on the fence here, but leaning to us keeping the status quo, this has been a very positive step and maybe we should pursue it for a while longer.
We have a very talented PTL and he need the help of the cores and the ‘decomposed’ plugins to all work together to address the issues. The community and stadium are evolving as we mature. Sometimes for the better and sometimes for the worse.
Putting aside the bike shedding one of the pain points at hand is the fact that the core neutron code is being neglected and not enough people are involved there. So maybe if we as a community would focus a little more on the core parts then we would not be having these discussions.
Thanks
Gary



From: "Armando M." <armamig at gmail.com<mailto:armamig at gmail.com>>
Reply-To: OpenStack List <openstack-dev at lists.openstack.org<mailto:openstack-dev at lists.openstack.org>>
Date: Friday, February 5, 2016 at 11:37 PM
To: OpenStack List <openstack-dev at lists.openstack.org<mailto:openstack-dev at lists.openstack.org>>
Cc: Ayal Baron <ayal.baron at huawei.com<mailto:ayal.baron at huawei.com>>, Eran Gampel <Eran.Gampel at huawei.com<mailto:Eran.Gampel at huawei.com>>
Subject: Re: [openstack-dev] [Neutron] Evolving the stadium concept



On 5 February 2016 at 05:41, Gal Sagie <gal.sagie at gmail.com<mailto:gal.sagie at gmail.com>> wrote:
Armando,

I think that contributing and innovating in Dragonflow to implement Neutron in an open way and serve as an  alternative and as an example
for distributed networking patterns IS driving Neutron forward, i am very sad that you fail to see this and try to pick to
my review/patches count.

Beside the big over head i devote to Dragonflow, due to the fact that it really runs as an open source project, i also help and contribute
as much as i can to OVN and of course my efforts in Kuryr, which to me solves a critical and important thing for Neutron and for OpenStack
in mixed containers environments.
(And the rest of the time that i try to devote to Neutron and other sub-projects, currently still under Neutron big-stadium)

Of course that all of this in addition to my efforts and success to convince and assist in bringing more people
and more companies to contribute in an open way with the community in many areas in Neutron (some you are familiar with like the border gateway and l2gateway others that you are not..),
both internally and externally, writing blogs/arranging meetups to promote and extend some
of the above projects visibility and Neutron as such.

Believe me that i truly am passionate about Neutron, OpenStack and open source and try my best to help and
contribute when ever i can and many times not due to my "Job requirement", i apologise that this is
not enough for you, there is only a limited amount of hours in a day :)

However, i truly believe that Dragonflow, and ANY other true open source implementation of Neutron helps move
Neutron forward and i hope to continue do so either as Neutron big-stadium, as a big-tent project or as something else.

I don't recall pointing at any stats. I appreciate your sales pitch, but that doesn't change the fact that when looking at features like port forwarding and tags (stuff that you indeed proposed and that can be beneficial for the project as a whole), you didn't seem to give them enough priority, meaning that Neutron core is not a priority for you...but don't get me wrong...everyone has his/her own priorities.

My point being contributing to Dragonflow/OVN/etc alone is great because it drives adoption and provide choice, but it is not enough to provide benefit and value to the rest of projects and initiatives that exist within the Neutron ecosystem, and use Neutron as a backbone to deliver network services.

There's a wealth of more or less glamorous activities that the core team is responsible for and are the true blood of this project. If the heart doesn't pump out this blood to the limbs, the limbs might as well die.


As i have talked with Russell and explained, to me the Big Stadium was/is a way to keep Networking related projects "near"
the group of people that has the best context to review / help and comment, its obviously not working and thats fine, lets
try something different...


On Thu, Feb 4, 2016 at 8:18 PM, Armando M. <armamig at gmail.com<mailto:armamig at gmail.com>> wrote:


On 4 February 2016 at 04:05, Gal Sagie <gal.sagie at gmail.com<mailto:gal.sagie at gmail.com>> wrote:
Hi Assaf,

I think that if we define a certain criteria we need to make sure that it applies to everyone equally.
and it is well understood.

I must admit I am still waking up and going through the entire logs etc. However I cannot help but point out that one criteria that Russell and other TC people are behind (me included) is the significant 'team overlap' (and I would add it to be for a prolonged amount of time). This doesn't mean just drop the accidental bug fix or enhancement to enable the subproject to work with Neutron or address the odd regression that sneaks in from time to time, but it means driving Neutron forward so that it is beneficial for the project as a whole.

If you look at yourself, can you candidly say that you are making an impact to the core of Neutron? You seem you have dropped off the radar in the Mitaka timeframe, and haven't made a lasting impact in the Liberty timeframe. I applaud your Kuryr initiative and your specs proposals, but both are not enough to warrant Dragonflow for inclusion.

If the team overlap changes, then great, we'll reassess.

That said, I'll continue my discussion on the patch...

I have contributed and still am to both OVN and Dragonflow and hope to continue do so in the future,
i want to see both of these solutions become a great production grade open source alternatives.

I have less experience in open source and in this community from most of you, but from what i saw users
do take these things into consideration, its hard for a new user and even not so new to understand the possibilities correctly
specially if we cant even define them ourselves

Instead of spending time on technology and on solving the problems for our users we are concentrating
on this conversation, we haven't even talked about production maturity, feature richness and stability as you say
and by doing this move, we are signaling something else for our users without actually discussing about all the
former ourselves.

I will be ok with what ever the Neutron team decide on this, as they can define the criteria as they please.
Just shared my opinion on this process and my disappointment from it as someone who values open source
a lot.

Gal.


On Thu, Feb 4, 2016 at 11:31 AM, Assaf Muller <amuller at redhat.com<mailto:amuller at redhat.com>> wrote:
On Thu, Feb 4, 2016 at 10:20 AM, Assaf Muller <amuller at redhat.com<mailto:amuller at redhat.com>> wrote:
> On Thu, Feb 4, 2016 at 8:33 AM, Gal Sagie <gal.sagie at gmail.com<mailto:gal.sagie at gmail.com>> wrote:
>> As i have commented on the patch i will also send this to the mailing list:
>>
>> I really dont see why Dragonflow is not part of this list, given the
>> criteria you listed.
>>
>> Dragonflow is fully developed under Neutron/OpenStack, no other
>> repositories. It is fully Open source and already have a community of people
>> contributing and interest from various different companies and OpenStack
>> deployers. (I can prepare the list of active contributions and of interested
>> parties) It also puts OpenStack Neutron APIs and use cases as first class
>> citizens and working on being an integral part of OpenStack.
>>
>> I agree that OVN needs to be part of the list, but you brought up this
>> criteria in regards to ODL, so: OVN like ODL is not only Neutron and
>> OpenStack and is even running/being implemented on a whole different
>> governance model and requirements to it.
>>
>> I think you also forgot to mention some other projects as well that are
>> fully open source with a vibrant and diverse community, i will let them
>> comment here by themselves.
>>
>> Frankly this approach disappoints me, I have honestly worked hard to make
>> Dragonflow fully visible and add and support open discussion and follow the
>> correct guidelines to work in a project. I think that Dragonflow community
>> has already few members from various companies and this is only going to
>> grow in the near future. (in addition to deployers that are considering it
>> as a solution)  we also welcome anyone that wants to join and be part of the
>> process to step in, we are very welcoming
>>
>> I also think that the correct way to do this is to actually add as reviewers
>> all lieutenants of the projects you are now removing from Neutron big
>> stadium and letting them comment.
>>
>> Gal.
>
> I understand you see 'Dragonflow being part of the Neutron stadium'
> and 'Dragonflow having high visibility' as tied together. I'm curious,
> from a practical perspective, how does being a part of the stadium
> give Dragonflow visibility? If it were not a part of the stadium and
> you had your own PTL etc, what specifically would change so that
> Dragonflow would be less visible. Currently I don't understand why
> being a part of the stadium is good or bad for a networking project,
> or why does it matter. Looking at Russell's patch, it's concerned with
> placing projects (e.g. ODL, OVN, Dragonflow) either in or out of the
> stadium and the criteria for doing so, I'm just asking how do you
> (Gal) perceive the practical effect of that decision.

Allow me to expand:
It seems to me like there is no significance to who is 'in or out'.
However, people, including potential customers, look at the list of
the Neutron stadium and deduce that project X is better than Y because
X is in but Y is out, and *that* in itself is the value of being in or
out, even though it has no meaning. Maybe we should explain what
exactly does it mean being in or out. It's just a governance decision,
it doesn't reflect in any way of the quality or appeal of a project
(For example some of the open source Neutron drivers out of the
stadium are much more mature, stable and feature full than other
drivers in the stadium).

>
>>
>>
>> On Wed, Feb 3, 2016 at 11:48 PM, Russell Bryant <rbryant at redhat.com<mailto:rbryant at redhat.com>> wrote:
>>>
>>> On 11/30/2015 07:56 PM, Armando M. wrote:
>>> > I would like to suggest that we evolve the structure of the Neutron
>>> > governance, so that most of the deliverables that are now part of the
>>> > Neutron stadium become standalone projects that are entirely
>>> > self-governed (they have their own core/release teams, etc).
>>>
>>> After thinking over the discussion in this thread for a while, I have
>>> started the following proposal to implement the stadium renovation that
>>> Armando originally proposed in this thread.
>>>
>>> https://review.openstack.org/#/c/275888
>>>
>>> --
>>> Russell Bryant
>>>
>>> __________________________________________________________________________
>>> OpenStack Development Mailing List (not for usage questions)
>>> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe<http://OpenStack-dev-request@lists.openstack.org?subject:unsubscribe>
>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
>>
>>
>>
>>
>> --
>> Best Regards ,
>>
>> The G.
>>
>> __________________________________________________________________________
>> OpenStack Development Mailing List (not for usage questions)
>> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe<http://OpenStack-dev-request@lists.openstack.org?subject:unsubscribe>
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
>>

__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe<http://OpenStack-dev-request@lists.openstack.org?subject:unsubscribe>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev



--
Best Regards ,

The G.

__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe<http://OpenStack-dev-request@lists.openstack.org?subject:unsubscribe>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev



__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe<http://OpenStack-dev-request@lists.openstack.org?subject:unsubscribe>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev




--
Best Regards ,

The G.

__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe<http://OpenStack-dev-request@lists.openstack.org?subject:unsubscribe>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20160207/cc31314d/attachment.html>


More information about the OpenStack-dev mailing list