Hi Etienne, all,
your points are interesting with thanks to some more information to Dan also.
Finally, the use cases in the document are very general. I personally would like to see some reasons why this should be an OGF standard, rather than it just being some software that is developed and then used by a few groups. Who are the users of this potential standard?
Also, on top of these questions, I also seek to understand why this is specifically an PGI topic?
VM/Application (repository) interoperability are similar aspects that we might ‘discuss’ within GIN first in order to understand the potential for standardization activities.
If you like, you can have a short talk in the GIN session at the OGF Next week (~10 minutes) shortly pointing out your answers to the aforementioned questions.
Please contact me off-line when you are interested so that we can plan this in the session.
In the context of PGI, I would like to focus on our given scope.
Take care,
Morris
Von: Daniel S. Katz [mailto:d.katz@ieee.org]
Gesendet: Dienstag, 5. Juli 2011 21:24
An: Etienne URBAH
Cc: Riedel, Morris; Johannes WATZL; pgi-wg@ogf.org; Ad EMMEN; edgi-na2@mail.edgi-project.eu; lodygens@lal.in2p3.fr; Jozsef KOVACS
Betreff: Re: [Pgi-wg] OGF PGI : Proposal to create an OGF recommendation for Application Repositories
Hi Etienne,
This is an interesting idea. A few quick thoughts follow:
You may also want to consider image (VM) repositories. As an example from FutureGrid, see: http://grids.ucs.indiana.edu/ptliupages/publications/vonLaszewski-draft-11-imagerepo.pdf I would ideally like this effort to consider if application repositories and image repositories might be subclasses of a more general repository.
Another thought is that applications here are implicitly code that is downloaded, rather than services. If services were also included, including SaaS components, what would change?
Overall, these two points are an effort to make OGF consider grids and clouds together, rather than separately.
Finally, the use cases in the document are very general. I personally would like to see some reasons why this should be an OGF standard, rather than it just being some software that is developed and then used by a few groups. Who are the users of this potential standard?
Thanks,
Dan
On Jul 5, 2011, at 1:58 PM, Etienne URBAH wrote:
Morris, Johannes and all,