Response from representatives of US Physics Grid Middleware to HEPCAL document R. Pordes, Sept 8th 0 - PowerPoint PPT Presentation

1 / 3
About This Presentation
Title:

Response from representatives of US Physics Grid Middleware to HEPCAL document R. Pordes, Sept 8th 0

Description:

Response from representatives of US Physics Grid Middleware to HEPCAL document ... Picky comments on some of the individual use cases were sent to WP8 and there is ... – PowerPoint PPT presentation

Number of Views:38
Avg rating:3.0/5.0
Slides: 4
Provided by: pc6114
Category:

less

Transcript and Presenter's Notes

Title: Response from representatives of US Physics Grid Middleware to HEPCAL document R. Pordes, Sept 8th 0


1
Response from representatives of US Physics Grid
Middleware to HEPCAL documentR. Pordes, Sept 8th
02
  • Pointer to the document distributed to PPDG and
    iVDGL collaborations.
  • Document read and discussed at US ATLAS Grid
    meeting and by US CMS SC management. Any
    comments fed back through the overall experiment
    representatives.
  • Solicited specific comments on the document from
    Ian Foster, Miron Livny, Ruth Pordes and the
    PPDG Analysis Tools workshop at LBL in June.

2
Comments ..
  • Document is a comprehensive and valuable record
    of requirements for grid data processing and
    analysis.
  • Some of the use cases  - especially the monte
    carlo and analysis jobs - are  still  high level
    and will need more analysis to understand the 
    design and implementations needed to deliver
    them.
  • The requirements will need to be revisited as
    more experience is gained in deploying and
    running the experiment applications.
  • The overall functionality, principles and
    usability of an integrated system(s) are not yet
    addressed .e.g scale of deployment, level of
    integration - source, binary, executables
    performance, reliability There is no mention of
    the need for event and notification services
  • There would be benefit from adding some level of
    requirements on error reporting, propagation, and
    consistency.

3
And..
  • We endorse the recommendation of the RTAG group
    in saying thatmore work is needed to understand
    the scope and nature ofcatalogs and their place
    in the functional hierachy
  • the analysis use cases would benefit from some
    drilling down to the next level of service
    components. More attention to the requirements
    for interactive analysis is needed.
  • Review of the requirements by representatives of
    a running experiment starting to do distributed
    analysis - e.g. BaBar, D0 - would be worthwhile.
  • Picky comments on some of the individual use
    cases were sent to WP8 and there is email
    exchange with Jeff Templon etc.
Write a Comment
User Comments (0)
About PowerShow.com