[ogsa-wg] Teleconference minutes - 19 October 2005

[Hiro Kishimoto]

Hi Andreas,

I joined the call 30 minutes late and missed the minutes approval.
Actually, I have a couple of comments on them and proposed to
fix them.

 >   - Teleconference minutes for Sep 28, Oct 12 and 17 approved with no
 >     changes

(1) Oct. 12, GGF15 and F2F evaluation;
Number of attendees is "50" instead of "15."

"GGF participant number where down" should read "GGF participant number
*were* down."

(2) Oct. 12, Next F2F;
Steven also ask me to schedule BES and info model sessions early in the
week (if possible).

 >   - GGF15 minutes approved with no changes
 >   - F2F minutes approved with no changes

(3) Participants for Oct. 6 meeting;
Affiliation of Wolfgang Ziegler should read "Fraunhofer."

(4) Review of Ellen's Approach slides;
In order to avoid possible confusion, I want to use term "CIM Profile"
instead of just "Profile."

Thanks,
----
Hiro Kishimoto

Andreas Savva wrote:
> Minutes attached.
> 
> https://forge.gridforum.org/projects/ogsa-wg/document/minutes-20051019/en/1
> 
> -- 
> Andreas Savva
> 
> 
> ------------------------------------------------------------------------
> 
> OGSA Teleconference - 19 October 2005
> =====================================
> 
> * Participants
> 
>   Mike Behrens (R2AD, LLC)
>   Andrew Grimshaw (UVa)
>   Hiro Kishimoto (Fujitsu)
>   Allen Luniewski (IBM)
>   Tom Maguire (EMC)
>   Takuya Mori (NEC)
>   Steven Newhouse (OMII)
>   Andreas Savva (Fujitsu)
>   Ellen Stokes (IBM)
>   Ravi Subramaniam (Intel)
>   Jay Unger (IBM)
>   Pete Ziu (Northrop Grumman)
> 
>   Minutes: Andreas Savva
> 
> * Minutes approval
> 
>   - Teleconference minutes for Sep 28, Oct 12 and 17 approved with no
>     changes
>   - GGF15 minutes approved with no changes
>   - F2F minutes approved with no changes
> 
> * Management Developement Conference presentation
> 
>   - Confirmed Ravi as the presenter
>   - There is an existing slide deck. Andrew will upload it to
>     Gridforge and the group will review it next week.
> 
>   - Hiro to add the review of this presentation on next Monday's
>     call. Scheduled time 20 minutes.
> 
> * EMS discussion
> 
>   - Review of UML sequence diagram previously sent out by Steven N.
>   - Background:
>     - Initially generated during last F2F and partly revised
>       afterwards.
>     - This is scenario 2.3
>     - Clarified that this is *a* scenario not *the* scenario to
>       address this particular stage of the EMS Roadmap. There may be
>       other sequences.
>   - The JSDL document is initially abstract and becomes more concrete,
>     for example, as a result of the EPS invocation.
> 
>   - The EMS Roadmap should also list the JSDL features that should be
>     supported and make clearer what is becoming more concrete.
> 
>   - Where is ACS? 
>     - ACS is part of a later scenario according to this Roadmap.
> 
>   - Why not include CSG in this diagram? 
>     - EPS is seen as a black box at this point. It may be calling CSG
>       behind the scenes but this interaction does not need to be shown
>       in this diagram.
>     - Why not rename EPS to RSS then? 
>       - RSS is the name of the group not of a service or interface.
> 
>   - The positioning of Agreement
>     - Donal pointed out on the list that the return of EPS may be some
>       form of Agreement.
>     - Details are not clear yet. May have to put together another
>       diagram to show this. Or it may be clearer in later stages of
>       the Roadmap, e.g., when including reservations in the sequences.
> 
>   - Ravi volunteered to refine some of these scenarios. He will
>     provide text for the EMS Roadmap.
> 
>   - Why is the job and its management not shown in the sequence diagram?
>     - Its representation and management are out of scope of this stage
>       of the Roadmap.
>     - Is a sequence diagram of stopping the job also out of scope at
>       this stage? Yes.
> 
>   - The EMS Architecture document should include more detail to
>     describe everything in the sequence diagram, e.g., the ordering
>     policy mentioned in the call to EPS.
>     - It is not the intention for the EMS Architecture document to be
>       normative but it should be more detailed than what's in OGSA
>       Architecture (1.0 or later). The OGSA Architecture document
>       should contain a higher level summary of what will (eventually)
>       be in the EMS Architecture document.
>     - Starting point: 
>       - The EMS section from OGSA 1.0 (or 1.5); and
>       - the older EMS draft (pre-OGSA 1.0) (Andreas has sent this to
>         Steven. It is somewhat dated but useful as background
>         information.)
> 
>   - OGSA-WG should agree on a common UML tool so as to share the work
>     of drawing up these sequence diagrams.
>     - Criteria: free for use and reasonable performance
>     - Andreas has a list of (free) tools and will send it out.
>     - JUDE is being used by some people. It comes in various versions
>       with the free versions having some limitations.
> 
> * Charter
> 
>   - Andrew to take the OGSA-WG charter to the GFSG for approval.
> 
> * Other business
> 
>   - Postponed the Information model discussion for next week.