CMS: a script to extract configurations for data taking

[katilp]

All configuration files used in data taking of CMS data are stored and made available as COPD records

The configurations are listed in

• https://fwyzard.web.cern.ch/fwyzard/hlt/2010/summary
• https://fwyzard.web.cern.ch/fwyzard/hlt/2011/summary
• https://fwyzard.web.cern.ch/fwyzard/hlt/2012/summary
• https://fwyzard.web.cern.ch/fwyzard/hlt/2015/summary
• https://fwyzard.web.cern.ch/fwyzard/hlt/2016/summary
• https://fwyzard.web.cern.ch/fwyzard/hlt/2017/summary
• https://fwyzard.web.cern.ch/fwyzard/hlt/2018/summary

Summary pages based on the information of listings above are provided for each year of data taking:

• 2010: http://opendata.cern.ch/record/1699
• 2011: http://opendata.cern.ch/record/1700
• 2012: http://opendata.cern.ch/record/1701

The configuration files have been extracted manually from https://cmsweb.cern.ch/confdb/

Write a script to extract the configuration files for Run2 data taking.

The CODP summary pages show only the data-taking periods which are open data even if all config files will be extracted. 2015 open data will be RunD, see #70 run numbers
Minimum: 256630
Maximum: 260627

[katilp]

A query example

python HLTrigger/Configuration/python/Tools/confdbOfflineConverter.py --configName /cdaq/physics/Run2015/25ns14e33/v3.3/HLT/V3 --adg > hltMenu_2015_v3p3.py
Need to be in a relatively recent release, ie >= 10_6_3 as the database hosts changed.

The script above can be run on lxplus after having done:

cmsrel CMSSW_10_6_10
cd CMSSW_10_6_10/src
cmsenv
git cms-addpkg HLTrigger/Configuration

For the naming of the files, use the following (example):
/cdaq/physics/firstCollisions15/v6.0/HLT/V1 -> cdaq_physics_firstCollisions15_v6.0_HLT_V1.py

[katilp]

@tiborsimko where would you like to have the config files stored for Run2 data-takings years? Should we create /eos/opendata/cms/configuration-files/2015 to start with (or use some temporary area)?

[tiborsimko]

Usually it is stored locally in data-curation caches, until the time we have the list of datasets and their provenance nailed down, and then it is copied over to the EOS location indeed.

[katilp]

This script (and what follows later in #74) can go in a new directory named cms-run2-hlt-triggers with a similar structure as cms-2011-hlt-triggers i.e. two subdirectories: code and inputs

Make your own fork of https://github.com/cernopendata/data-curation to do this

[mokotus]

Will be integrated together with #74, closing for now