[ClusterLabs Developers] [HA] future of OpenStack OCF resource agents (was: resource-agents v4.2.0)

[Adam Spiers]

Kristoffer Grönlund <kgronlund at suse.de> wrote:
>On Wed, 2018-10-24 at 13:25 +0100, Adam Spiers wrote:
>> [cross-posting to openstack-dev]
>>
>> Oyvind Albrigtsen <oalbrigt at redhat.com> wrote:
>> [snipped]
>>
>> > - openstack-cinder-volume
>> > - openstack-floating-ip
>> > - openstack-info
>>
>> That's an interesting development.
>>
>> By popular demand from the community, in Oct 2015 the canonical
>> location for OpenStack-specific resource agents became:
>>
>>     https://git.openstack.org/cgit/openstack/openstack-resource-agent
>> s/
>>
>> as announced here:
>>
>>     http://lists.openstack.org/pipermail/openstack-dev/2015-October/0
>> 77601.html
>>
>
>Hi Adam,
>
>I guess I did know that but it didn't cross my mind to question the
>submission, apologies. What is the relationship of the submitter of
>these openstack agents to the upstream project?

I don't know.

>Are these agents copies
>of your openstack agents, or completely separate developments?

Completely separate developments.

Ironically, it seems that he submitted a blueprint for this work
to openstack-resource-agents back in May, but somehow I totally
missed it:

    https://blueprints.launchpad.net/openstack-resource-agents/+spec/resource-agents-consuming-openstack-api

>Maybe the best solution would be if you could get in touch with the
>submitter of these agents and see if you can join forces in maintaining
>a single set of agents.

Isn't that what I just did, by cross-posting to both communities? ;-)

Unfortunately I don't have a direct email address I can Cc.

>I guess this is an issue with maintaining agents outside the resource-
>agents repository - the same goes for the two PostgreSQL agents in
>existence. It's possible that people don't even realise that there
>could be agents that are not maintained in the resource-agents
>repository. :/ Maybe one solution would be to add documents in place of
>the agents with links to their actual upstream.

Right.  Discoverability is key.