TOTCSI User Guide

This document contains a brief summary of steps which should be undertaken in order to execute a reconstruction or simulation process using CMSSW 7.0.4 and TOTCSI 2.2 software.

Requirements

• At least 700 MB of free space on AFS file system for TOTCSI workspace containing configuration files and logs.
• A sufficiently large directory for output files on one of the following file systems: AFS, CASTOR or EOS.
• Jobs have to be submitted to lxbatch cluster from lxplus machines.

Reconstruction

In the directory "mydir_for_reco" do fthe following

• Copy example reconstruction configuration from AFS.

cp -R /afs/cern.ch/exp/totem/soft/TOTCSI/examples/scenarios/rec01 .

• Addjust following parameters in the file totcsi_rec01.py:
  • config.parallel1_output - path on CASTOR where raw output files will be stored
  • config.sequential_output - path on CASTOR where ntuple file will be stored
  • config.reconstruction.input_data - run number
  • config.reconstruction.events_per_file_to_reconstruct - number of events to be reconstructed using input file
  • config.bsub.queue - queue type, which determines how long given job can be processed (“1nh” - 1h queue, “8nh” - 8h queue, “1nd” - 1 day queue, “1nw” - 1 week queue, ...)
• Clean both workspace and output directory on CASTOR.

rm -rf workspace
nsrm -r <path to parallel1_output directory>
nsrm -r <path to sequential_output directory>

• Read recursively content of filesystem_base_dir in search for input files.

/afs/cern.ch/exp/totem/soft/TOTCSI/totcsi -c ./totcsi_rec01.py rf

• Choose input files associated with the run number.

/afs/cern.ch/exp/totem/soft/TOTCSI/totcsi -c ./totcsi_rec01.py fp

• Generate necessary files and directories for job submission.

/afs/cern.ch/exp/totem/soft/TOTCSI/totcsi -c ./totcsi_rec01.py gs

• Submit jobs creating raw output files.

bash workspace/totcsi_rec01/main_submit_parallel1.sh

• Wait until all jobs are done. You can control job progress using bjobs command.
• Check whether raw output files were generated.

/afs/cern.ch/exp/totem/soft/TOTCSI/totcsi -c ./totcsi_rec01.py cp1

• Submit job creating ntuple file.

bash workspace/totcsi_rec01/main_submit_sequential.sh

• Wait until this job is done. Again you can control job progress via bjobs command.
• Check whether ntuple file was generated.

/afs/cern.ch/exp/totem/soft/TOTCSI/totcsi -c ./totcsi_rec01.py cs

Simulation

• Copy example simulation configuration from AFS.

cp -R /afs/cern.ch/exp/totem/soft/TOTCSI/examples/scenarios/sim01 .

• Addjust following parameters in the file totcsi_sim01.py:
  • config.parallel1_output - path on CASTOR where raw output files will be stored
  • config.sequential_output - path on CASTOR where ntuple file will be stored
  • config.simulation.number_of_events - number of events per job
  • config.simulation.number_of_jobs - number of job to be created
  • config.bsub.queue - queue type, which determines how long given job can be processed (“1nh” - 1h queue, “8nh” - 8h queue, “1nd” - 1 day queue, “1nw” - 1 week queue, ...)
• Clean both workspace and output directory on CASTOR.

rm -rf workspace
nsrm -r <path to parallel1_output directory>
nsrm -r <path to sequential_output directory>

• Generate necessary files and directories for job submission.

/afs/cern.ch/exp/totem/soft/TOTCSI/totcsi -c ./totcsi_sim01.py gs

• Submit jobs creating raw output files.

bash workspace/totcsi_sim01/main_submit_parallel1.sh

• Wait until all jobs are done. You can control job progress using bjobs command.
• Check whether raw output files were generated.

/afs/cern.ch/exp/totem/soft/TOTCSI/totcsi -c ./totcsi_sim01.py cp1

• Submit job creating ntuple file.

bash workspace/totcsi_sim01/main_submit_sequential.sh

• Wait until this job is done. Again you can control job progress via bjobs command.
• Check whether ntuple file was generated.

/afs/cern.ch/exp/totem/soft/TOTCSI/totcsi -c ./totcsi_sim01.py cs

Advanced configuration

For reconstruction you can also adjust following parameters in totcsi_rec01.py configuration file:

• config.cmssw_dir - path to the CMSSW release (doesn’t need to be changed if you want to use official, latest version)
• config.filesystem_base_dir - path to the directory containing raw input data
• config.input_file_name_extensions - list of raw input data files extensions which should be processed, i.e. “vmeb”

For simulation you can also adjust following parameters in totcsi_sim01.py configuration file:

• config.cmssw_dir - path to the CMSSW release (doesn’t need to be changed if you want to use official, latest version)

For more information please read README.txt files associated with examples. You can find them here:

• /afs/cern.ch/exp/totem/soft/TOTCSI/examples/scenarios/rec01/README.txt
• /afs/cern.ch/exp/totem/soft/TOTCSI/examples/scenarios/sim01/README.txt

Execution time

TOTCSI version 2.0

This obsolete version runs on SLC5 machines (at the end of 2014 lxplus nodes with SLC5 were migrated to SLC6). It supports CMSSW versions compatible with SLC5: CMSSW_4_2_4 and earlier.

TOTCSI developers

For more information on project development please visit TOTCSI Developer Guide.