[openstack-dev] [heat] Questions on template-validate

Jay Dobies jason.dobies at redhat.com
Wed Feb 24 14:13:28 UTC 2016



On 02/24/2016 02:18 AM, Anant Patil wrote:
> On 23-Feb-16 20:34, Jay Dobies wrote:
>> I am going to bring this up in the team meeting tomorrow, but I figured
>> I'd send it out here as well. Rather than retype the issue, please look at:
>>
>> https://bugs.launchpad.net/heat/+bug/1548856
>>
>> My question is what the desired behavior of template-validate should be,
>> at least from a historical standpoint of what we've honored in the past.
>> Before I propose/implement a fix, I want to make sure I'm not violating
>> any unwritten expectations on how it should work.
>>
>> On a related note -- and this is going to sound really stupid that I
>> don't know this answer -- but are there any docs on actually using Heat?
>> I was looking for docs that may explain what the expectation of
>> template-validate was but I couldn't really find any.
>>
>> The wiki links to a number of developer-centric docs (HOT guide,
>> developer process, etc.). I found the (what I believe to be current)
>> REST API docs [1] but the only real description is "Validates a template."
>>
>> Thanks  :D
>>
>>
>> [1] http://developer.openstack.org/api-ref-orchestration-v1.html
>>
>> __________________________________________________________________________
>> 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
>>
>
> Sometime back, I too went through this, but got adjusted to the thought
> that the template validation is really for validating the syntax and
> structure of a template. Whether the values provided are valid or not
> will be decided when the stack is validated. The values that depend on
> resource plugins to fetch data from other services are not validated,
> and to me it makes sense. It helps user to quickly test-develop
> templates that are syntactically and structurally valid and they don't
> have to depend on resource plugins and services availability. IMO, it
> would be better to document the way template validate works, than to
> make it a heavy weight request that depends on plugins and services.

Everything you're saying makes sense. I like the idea of it as a syntax 
validation of the structure of the template alone. I also like that it's 
lightweight.

My only concern is the inclusion of the value in the returned template. 
That's the part that feels weird to me, and is especially misleading if 
we don't have docs around it.

I'm with you on the idea of flushing out those docs. I'll ask in the 
meeting today on the best way to pursue that. I know I've seen patches 
related to updating the docs in code for resource plugins, but I'm not 
sure if that covers the external API docs. If it doesn't, I'll file a 
blueprint for that so we can track it as an across-the-board API docs 
enhancement.

> __________________________________________________________________________
> 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