Data Challenge 2

1
Data Challenge 2 Towards a Work Plan Richard
Dubois
2
Work Areas

• Pipeline/Data Server/Display Server
• Instrument Simulation/Reconstruction
  GlastRelease
• Orbit simulation
• Sky Model sources and backgrounds
• Science Tools

3
Pipeline Data Servers

• Pipeline
• In test now for Flight Int EM2 coming soon for
  first real use
• Not quite ready for bulk MC (limit on db
  connections fix in progress)
• Expect Flight Int developments will be applicable
  to DC2
• Data Server
• Navid's DC1 server needs to learn the pipeline
  dataset catalogue for access
• Need to gather up interested folks and spec out
  DC2 version
• Options can be SSC DC1 copy SSC DC1
  fully integrated L1 L0.5
• SLAC Computing Services interested in helping
• Mega Memory server in the works (9 TB in coming
  year aiming for 250 TB RAM!)

4
Display Servers

• Handy for instrument studies
• Perhaps only way to allow science community to
  view events in detail
• FRED can operate in remote mode
• Hook up local display to remote server
• In principle, just set up a bunch of boxes
  running Gleam in FRED-mode and allow users to
  hook up to them
• Need several new bits
• Corba connection server (hook up least-loaded
  server)
• Gleam able to find events in the dataset
  catalogue
• Ability to control server's access to events,
  without restarting Gleam
• Convenient control within FRED
• Note this is NOT a web plugin

5
Servers Schedule
6
Sim/Recon

• Instrument Level stuff
• Upgrade TkrRecon, CalRecon
• Allow different particle hypotheses
• Event Pre-Filter
• Cal Clustering, MIP finding, improved Energy
  corrections
• Redo background analysis long job!
• Regenerate IRFs
• Need lead time for SciTools to at least know
  functional form

7
GR Schedule
8
Source Orbit Upgrades

• Sky Model
• Interstellar emission model
• Phony ephemerides for pulars
• Variable sources AGNs and pulsars, including
  binaries
• Test well ahead of event generation!
• Orbit sim
• More realistic than DC1
• eg non-infinite slew rates

9
Sources/Orbit Schedule
10
Science Tools

• Likelihood
• Binned vs unbinned evaluations
• Zenith angle cuts
• Needs form of IRF sooner than last time
• GRB
• Evaluation of Xspec 12
• Unbinned analysis
• Spectral /or Temporal analysis
• Pulsars
• Period search
• Map Generation
• Source Catalogue
• Source detection
• Pipeline operation

11
Science Tools Schedule
12
Infrastructure Documentation

• Code distributions
• Linux installer done needs user testing
• Windows installer
• External libs nearly done
• The rest awaits Release Manager doing windows
  builds
• Documentation
• Tech writer (Chuck Patterson) hired to create
  User Workbook
• 3 months into the 1 year projected project
• Science Tools to be addressed after Flight
  Integration related doc (obvious timing)

13
Documentation Project

• User workbook is in progress
• Focusing on event analysis portion first to
  accommodate Flight Integration schedule
• Use of ntuples, RootTreeAnalysis macro and
  userAlg coming soon

14
Summary

• DC2 planning getting underway
• Based on DC1 experience
• Mostly more DC1
• ToDo list will evolve with discussions in the
  steering committee and collaboration
• Wild card is Flight Integration support
• Hardware delays push SAS support further into DC2
  prep period
• Lots of details to cover not mentioned here
• The price is right for OpenOffice, but PowerPoint
  is still better