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

[Philippe Merle]

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.'

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

You should.

>   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.


>   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.

>
> 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 ;-)

A+
Philippe

>
>> 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
>>