[openstack-dev] [kolla][TripleO] New areas for collaboration between Kolla and TripleO

Flavio Percoco flavio at redhat.com
Thu Jul 20 15:25:34 UTC 2017

On 20/07/17 08:18 -0700, Emilien Macchi wrote:
>On Thu, Jul 20, 2017 at 7:27 AM, Andy McCrae <andy.mccrae at gmail.com> wrote:
>> Hopefully that is useful, happy to discuss this more (or any other
>> collaboration points!) if that does sound interesting.
>> Andy
>> [1]
>> https://github.com/openstack/openstack-ansible-plugins/blob/master/action/_v2_config_template.py
>> [2]
>> https://docs.openstack.org/project-deploy-guide/openstack-ansible/draft/app-advanced-config-override.html
>Yes, this is very useful and this is what I also wanted to investigate
>more back in June:
>Like Flavio said, it sounds like we might just re-use what you guys
>did, since it looks flexible.
>What Doug wrote [1] stays very useful, since we don't want to re-use
>your templates, we would rather generate the list of options available
>in OpenStack projects by using oslo-.config directly. We could provide
>an YAML with key/values of things we want to generate in an inifile.
>Now we could ask ourselves, in that case, why not directly making
>oslo.config reading YAML instead of ini? Do we really need a
>User input → YAML →OSA config template plugin →INI →read by oslo.config
>we could have:
>User input → YAML → read by oslo.config
>I've discussed with some operators about this options but I want to
>re-iterate on it here. Any thoughts?
>[1] https://github.com/dhellmann/oslo-config-ansible

The plugin, as is, is capable of generating INI files as well as YAML files. I
don't really see the need of making oslo.config read YAML.

On one side the idea sounds appealing because well, we can do more with YAMl
than we can with INI files. On the other side, though, I don't think it's worth
the time right now.

A migration to YAML files requires way more work than we can account for. Not
only we need to make oslo.config support it, we also have to maintain
compatibility for quite a few cycles, etc.

Don't get me wrong. If someone wants to work on this, I'm good. What I'm saying
is that, at this point, I don't think it's worth it. It may be in the future.
The reality is that we depend on INI files now and we will for the forseable
future so the work has to be done anyway.


Flavio Percoco
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 862 bytes
Desc: not available
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20170720/bdba716e/attachment.sig>

More information about the OpenStack-dev mailing list