Difference between revisions of "Data Challenge 2: Ideas"

From GlueXWiki
Jump to: navigation, search
(Created page with "=New software= * BCAL reconstruction * CCDB * latest tracking * ROOT trees? * improved REST? =System Improvements= * farm management system at JLab ** crude set of perl script...")
 
(add in comments made at May 1 offline meeting)
 
(4 intermediate revisions by the same user not shown)
Line 1: Line 1:
 +
Thoughts on what goals we might have for the next data challenge.
 +
 
=New software=
 
=New software=
  
Line 6: Line 8:
 
* ROOT trees?
 
* ROOT trees?
 
* improved REST?
 
* improved REST?
 +
 +
=Simulation Configuration Changes=
 +
 +
* add electromagnetic background
 +
* add in dark noise in the BCAL
 +
* open up photon window, especially to higher photon energy, perhaps 7.5 GeV to the end-point
  
 
=System Improvements=
 
=System Improvements=
  
 +
* improve job success efficiency
 
* farm management system at JLab
 
* farm management system at JLab
 
** crude set of perl scripts now, full of kludges
 
** crude set of perl scripts now, full of kludges
Line 20: Line 29:
 
* produce simulated data as a completely separate step
 
* produce simulated data as a completely separate step
 
* write to tape library
 
* write to tape library
 +
* inefficient but realistic
 
* "raw data" sources:
 
* "raw data" sources:
 
*# JLab farm
 
*# JLab farm
Line 28: Line 38:
 
** Grid to JLab
 
** Grid to JLab
 
** JLab to Grid
 
** JLab to Grid
 +
 +
 +
----
 +
  
 
--[[User:Marki|marki]] 13:09, 1 May 2013 (EDT)
 
--[[User:Marki|marki]] 13:09, 1 May 2013 (EDT)

Latest revision as of 09:38, 15 May 2013

Thoughts on what goals we might have for the next data challenge.

New software

  • BCAL reconstruction
  • CCDB
  • latest tracking
  • ROOT trees?
  • improved REST?

Simulation Configuration Changes

  • add electromagnetic background
  • add in dark noise in the BCAL
  • open up photon window, especially to higher photon energy, perhaps 7.5 GeV to the end-point

System Improvements

  • improve job success efficiency
  • farm management system at JLab
    • crude set of perl scripts now, full of kludges
    • needs generalization, perhaps re-write
  • input and output data cataloging and management
    • database application?
    • grid-based meta-data catalog?

Raw Data Paradigm

  • produce simulated data as a completely separate step
  • write to tape library
  • inefficient but realistic
  • "raw data" sources:
    1. JLab farm
    2. online cluster
    3. grid
  • stresses different data paths
    • tape to farm at JLab
    • Grid to JLab
    • JLab to Grid




--marki 13:09, 1 May 2013 (EDT)