[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Charter goals
One idea we discussed in the call was (1) to define the JSIM and (2) to
identify other IMs that are needed from a Grid perspective, but to leave
the actual definition of these other IMs either to subsequent new WGs or to
this WG as future work under an extended scope. We may well tackle this in
a sequential way as Dan suggested.
I like this idea because that way, the goals for this WG are relatively
clearly defined, and it also allows for some flexibility to extend the
scope (then again quite clearly) once we have gained more experience in
defining such IMs.
I have one more comment: Tom mentioned OGSA as the driver for IMs, I think
we should reword that to be "the Grid community" or something like that,
because the requirements should be driven by the needs of the grid
community, not by the needs of OGSA.
In order to show that what we are doing has a strong relationship to OGSA,
I think we should mention that in addition to a CIM representation of the
IMs, we would also provide an interface definition of OGSA web services
(i.e. GSDL+XSD) according to a general mapping from CIM to OGSA that is
being defined in one of the OGSA related WGs.
Andy
Andreas Maier
IBM Senior Technical Staff Member, eServer Software Design
IBM Lab Boeblingen, maiera@de.ibm.com, +49-7031-16-3654
----- Forwarded by Andreas Maier/Germany/IBM on 25.09.2002 14:40 -----
Dan Gunter
<dkgunter@lbl.gov To: cgs-wg@gridforum.org
> cc:
Sent by: Subject: Re: Charter goals
owner-cgs-wg@grid
forum.org
24.09.2002 23:03
Well, who knows really. Neither of us have much to base our projections
on. Let's take a different approach. The group should first do the
queueing schema. Then if there's time, tackle OGSA. How does that sound?
So when the milestones are laid out they would go:
1. develop JSIM (6 mos. we hope)
2. address CIM/OGSA interactions (6-12 mos. we hope)
This way the group can focus on one thing at a time.
Of course the steering group might still say the charter is too broad,
but that's another matter.
-- Dan
Larry Flon wrote:
> IMHO, it's OK to set an 18mo time frame to address deficiencies in CIM as
> far as OGSA is concerned, and set the primary goal of this WG to be the
> JSIM. We should be able to deliver the JSIM in 6 mos, and we can
indentify
> the other key deficiencies of CIM, like no current support for TTL data,
> that need to be addressed before the WG disbands.
>
> Larry
>
> -----Original Message-----
> From: Dan Gunter [mailto:dkgunter@lbl.gov]
> Sent: Tuesday, September 24, 2002 1:35 PM
> To: cgs-wg@gridforum.org
> Subject: Re: Charter goals
>
>
> whoops! I meant that promoting things should NOT be an explicit task.
>
> makes more sense now, I hope :)
>
> - Dan
>
> Dan Gunter wrote:
>
>
>>Hi all,
>>
>>I missed the call, so I apologize if these points were already raised.
>>
>>It strikes me that the "primary goal" stated below is quite enough for
>>a
>>working group, and that the other "goals" are really just context and
>>future work that could be part of the charter's text, but really don't
>>belong as official WG goals.
>>
>>Also, I don't think "promoting" things should be an explicit task.
>>Useful WG products will promote themselves. Promoting something before
>>it's determined to be useful is not productive.
>>
>>- Dan
>>
>>Tom Roney wrote:
>>
>>
>>>One of the action items resulting from today's teleconference was the
>>>re-writing of the "Goals" section of our proposed "Charter of the CIM
>>>based Grid Schema Working Group." I volunteered to offer my
>>>interpretation of the changes we agreed to make. Your comments as to
>>>what I have perceived as a general concensus are quite welcome, and
>>>expected.
>>>
>>>You will note that we have agreed to rephrase the term
>>>"Job Submission Interface," in favor of "Job Submission Information
>>>Model." We are not describing an interface. Rather, we are
>>>describing the information required by an OGSA queueing service for
>>>the successful submission of a job to a queue.
>>>
>>> - The general goal of the CGS-WG is to produce CIM specifications
>>> and schemas of classes not yet standardized by the DMTF but
>>> urgently needed by the OGSA, and to promote these specifications
>>> and schemas within the DMTF for their adoption as CIM standards.
>>>
>>> - The primary goal of the CGS-WG is to produce a CIM-based
>>> "Job Submission Information Model" based on OGSA requirements,
>>> and to promote this model within the DMTF for their adoption
>>> as a CIM standard.
>>>
>>> - The secondary goal of the CGS-WG is to determine which other
>>> grid-computing resources are required by the OGSA but still
>>> lack a CIM specification and schema -- for development in
>>> either this or perhaps a future working group.
>>>
>>> - The CGS-WG will explore the possible use of evolving CIM-based
>>> Open Source software in grid computing, and liaise with other
>>> OGSA-based working groups for their input.
>>>
>>>Please comment,
>>>
>>>Tom
>>>
>>>