20–22 Jun 2005
Brno, Czech Republic
Europe/Zurich timezone

Session

WMS & DM & LB I

21 Jun 2005, 15:00
Brno, Czech Republic

Brno, Czech Republic

Description

This session should include:

  • Combining compute and data jobs in DAGs
  • WMS using data mgmt for
    • providing input data at execution site
    • storing results (sandboxes?)
  • Use of L&B for logging computational and data jobs
  • Adding job statistics (CPU time etc.) to L&B

Presentation materials

There are no materials yet.

  1. Fabrizio Pacini (Datamat SpA)
    21/06/2005, 15:00
    The current support for the handling of 'sparse' sandboxes via the WMS web service interface (a.k.a. the WMproxy) is described, and possibly demonstrated. Missing features and scenarios for extending this model to storage elements that do not provide a gridftp interface will be discussed.
    Go to contribution page
  2. Peter Kunszt (CERN)
    21/06/2005, 16:30
    There are many possibilities how to integrate 'data' and 'cpu' jobs. The CONDOR Stork component is one very good candidate to integrate to achive easy interoperability which would need to be explored.
    Go to contribution page
  3. Enzo Martelli & Francesco Prelz (INFN Milano)
    21/06/2005, 17:15
    We try to sort out a few details of which information needs to be passed among the various nodes of a DAG with data placement nodes, with a special care on when and how matchmaking should be performed.
    Go to contribution page
  4. Tiziana Ferrari, Elisabetta Ronchieri
    21/06/2005, 17:45
Building timetable...