Open Stack

Hello folks,

Just an FYI that I've submitted a pull request [1] to replace Celery with oslo.messaging.

I've tagged it as a work in progress per this note:

"Please review this CR, which replaces Celery with oslo.messaging components. I've verified that this works in my local environment, but I still need to add unit testing. I also need to verify that it works correctly with an HA Rabbit MQ cluster, as that is a hard requirement for Barbican."

Special thanks to Mark McLoughlin and Sylvain Bauza for pointing me to very useful links here [2] and here [3] respectively.

[1] https://review.openstack.org/#/c/60427/
[2] https://review.openstack.org/#/c/39929
[3] https://review.openstack.org/#/c/57880

Thanks,
John

________________________________________
From: Monty Taylor [mordred at inaugust.com]
Sent: Thursday, December 05, 2013 8:35 PM
To: Mark McLoughlin; Douglas Mendizabal
Cc: OpenStack Development Mailing List (not for usage questions); openstack-tc at lists.openstack.org; barbican at lists.rackspace.com
Subject: Re: [openstack-tc] [openstack-dev] Incubation Request for Barbican

On 12/06/2013 01:53 AM, Mark McLoughlin wrote:
> On Thu, 2013-12-05 at 23:37 +0000, Douglas Mendizabal wrote:
>>>
>>> I agree that this is concerning. And that what's concerning isn't so
>>> much that the project did something different, but rather that choice
>>> was apparently made because the project thought it was perfectly fine
>>> for them to ignore what other OpenStack projects do and go off and do
>>> its own thing.
>>>
>>> We can't make this growth in the number of OpenStack projects work if
>>> each project goes off randomly and does its own thing without any
>>> concern for the difficulties that creates.
>>>
>>> Mark.
>>
>> Hi Mark,
>>
>> You may have missed it, but barbican has added a blueprint to change our
>> queue to use oslo.messaging [1]
>>
>> I just wanted to clarify that we didn’t choose Celery because we thought
>> that “it was perfectly fine to ignore what other OpenStack projects do”.
>> Incubation has been one of our goals since the project began.  If you’ve
>> taken the time to look at our code, you’ve seen that we have been using
>> oslo.config this whole time.  We chose Celery because it was
>>
>> a) Properly packaged like any other python library, so we could just
>> pip-install it.
>> b) Well documented
>> c) Well tested in production environments
>>
>> At the time none of those were true for oslo.messaging.  In fact,
>> oslo.messgaging still cannot be pip-installed as of today.  Obviously, had
>> we know that using oslo.messaging is hard requirement in advance, we would
>> have chosen it despite its poor distribution story.
>
> I do sympathise, but it's also true is that all other projects were
> using the oslo-incubator RPC code at the time you chose Celery.
>
> I think all the verbiage in this thread about celery is just to
> reinforce that we need to be very sure that new projects feel a
> responsibility to fit closely in with the rest of OpenStack. It's not
> about technical requirements so much as social responsibility.
>
> But look - I think you've reacted well to the concern and hopefully if
> it feels like there was an overreaction that you can understand the
> broader thing we're trying to get at here.

I agree. I think you've done an excellent job in responding to it - and
I appreciate that. We're trying to be clearer about expectations moving
forward, which I hope this thread in some part helps with.

Monty