[occi-wg] Draft of OCCI 1.1 profile for VM templates/flavours

[Boris Parak]

On Thu, Mar 19, 2015 at 5:17 PM, Philippe Merle <philippe.merle at inria.fr> wrote:
> Hello Boris,
>
> Le 19/03/15 10:32, Boris Parak a écrit :
>>
>> Hello Philippe,
>>
>> On Thu, Mar 19, 2015 at 1:00 AM, Philippe Merle <philippe.merle at inria.fr>
>> wrote:
>>>
>>> Hello Boris,
>>>
>>> I have one comment on your OCCI Compute Resource Templates.
>>> This specification does not explicitly define in which category (Kind or
>>> Mixin) the attribute eu.egi.fedcloud.occi.compute.ephemeral is defined.
>>> It could be defined in each of the six mixins or as I did it could be
>>> defined in an ephemeral mixin inherited by the six mixins.
>>> I prefer the second solution, which factorizes common elements of the six
>>> mixins.
>>
>> it might not be explicitly mentioned there, but we adopted the first
>> approach you mentioned. We had a few reasons for doing that:
>>
>> 1.) the whole point of that attribute is setting a default value
>> different for each of those six mixins ; therefore it would have to be
>> "included" in all of them anyway ... and in a separate additional
>> mixin if we adopted the second approach (6 vs. 7 mixins with that
>> attribute)
>
>
> ok. I just pointed a drawback in the proposed specification.
> This specification defines an attribute but does not define in which
> category(ies) it is defined.
> So perhaps you could add a sentence to say 'This attribute is defined in
> each of the following six mixins.'

Good point. Will do.

>>
>> 2.) we do not use OCCI for modeling purposes,
>
>
> You should.

I know :-)

>>   features such as
>> attribute inheritance from related mixins are not implemented ;
>
>
> Do you want to say that
>
> attribute inheritance from related mixins are not implemented
>
>
> in the specific OCCI runtime you use.
>
> The OCCI core specification allows this.

Yes, specific runtime(s)/implementation(s).

>
>>   this
>> means that defining an attribute in one mixin and setting a default
>> value for that attribute in a different related mixin wouldn't work
>
>
> in the specific OCCI runtime you use, not in OCCI in general.

Exactly.

>
>>
>> 3.) this is all about practical applications for OCCI Infrastructure ;
>> we made it as simple as possible
>>
>> If that makes any sense to you ...
>
>
> In OCCIware we have industrial partners that do practical applications of
> OCCI.
> OCCIware is not just a pure research project. So this makes sense for me ;-)

Thank you for your feedback!

> A+
> Philippe
>

Cheers, Boris

>>
>>> A+
>>> Philippe
>>
>> Cheers, Boris
>>
>>> Le 17/03/15 13:21, Boris Parak a écrit :
>>>
>>> Hello everyone,
>>>
>>> I have here, for your consideration, one of the first "formal" outputs
>>> of EGI FedCloud's work with OCCI -- The OCCI Resource Template
>>> Profile. We would very much like to hear your opinions, comments
>>> and/or suggestions.
>>>
>>> After this (brief, hopefully) internal discussion phase, we would like
>>> to push for a public comment phase as soon as possible.
>>>
>>> Thank you!
>>>
>>> Cheers, Boris
>>> ---
>>> CESNET / EGI FedCloud
>>>
>>>
>>>
>>> _______________________________________________
>>> occi-wg mailing list
>>> occi-wg at ogf.org
>>> https://www.ogf.org/mailman/listinfo/occi-wg
>>>
>>>
>>>
>>> _______________________________________________
>>> occi-wg mailing list
>>> occi-wg at ogf.org
>>> https://www.ogf.org/mailman/listinfo/occi-wg
>>>
>