Skip to content

Latest commit

 

History

History
148 lines (115 loc) · 5.31 KB

README.md

File metadata and controls

148 lines (115 loc) · 5.31 KB

Embedding Configurations

This package collects the configs (cmsRun, gridcontrol, inputs dbs files) for embedding, such that one can start a large scale production

Current Status

Code Portation:

  • UL2016preVFP
  • UL2016postVFP
  • UL2017
  • UL2018

Additional features needed

  • Fiellist creation for completed Embedding Datasets
  • Dataset Publication
  • Integrate Dataset Publication into the main script
  • Extention to other computer infrastructures than KIT

UL Campaign

The setup is done automatically, using the create_UL_campaign.py tool.

usage: create_UL_campaign.py [-h] [--workdir WORKDIR] --era {2017,2018}
                             [--final-state {MuTau,ElTau,ElMu,TauTau,MuEmb,ElEmb}]
                             [--run RUN [RUN ...]] --mode
                             {preselection,full,nanoaod} --task
                             {setup_cmssw,upload_tarballs,setup_jobs,run_production,create_filelist,publish_dataset}
                             [--backend {etp,naf,lxplus}]
                             [--custom-configdir CUSTOM_CONFIGDIR] [--mc]
                             [--no_tmux]

Setup Grid Control for Embedding Production

optional arguments:
  -h, --help            show this help message and exit
  --workdir WORKDIR     path to the workdir
  --era {2017,2018}     Era used for the production
  --final-state {MuTau,ElTau,ElMu,TauTau,MuEmb,ElEmb}
                        Name the final state you want to process
  --run RUN [RUN ...]   Name or list of the runs you want to process, use all
                        to process all runs of an era
  --mode {preselection,full,nanoaod}
                        Select preselection mode, full embedding mode or
                        nanoaod mode
  --task {setup_cmssw,upload_tarballs,setup_jobs,run_production,create_filelist,publish_dataset}
                        Different commands that are possible
  --backend {etp,naf,lxplus}
                        Select the condor backend that is used.
  --custom-configdir CUSTOM_CONFIGDIR
                        If this is set, use the configdir from the given
                        folder
  --mc                  If this is set, mc embedding is run instead of data
                        embedding
  --no_tmux             If this is set, no tmux is used to run the jobs

Setup

Install the framework using

git clone --recursive git@github.com:KIT-CMS/gc_configs_for_embedding.git

Configurations

The large part of the embedding specific configuration settings can be found in scripts/ul_config.yaml

Preselection

Setup

For the preselection, only a single CMSSW version is needed. The version can be installed using

python3 create_UL_campaign.py --mode preselection --era 2018 --task setup_cmssw --run all

then, the different preselection tasks for all runs in a single era can be setup using

python3 create_UL_campaign.py --mode preselection --era 2018 --task setup_jobs --workdir /path/to/workdir --run all

or a single run can be specified by using the name of the run instead of all. If a space sepatated list of Runs is provided, those Runs will be processed. The workdir is the folder, that grid-control will use to keep track of the different jobs and store the respective job logfiles.

Production

The Production of the preselection can be started using

python3 create_UL_campaign.py --mode preselection --era 2018 --task run_production --run all

This will automatically start the fitting grid control tasks.

Output Collection

After successful completion of the preselection task, the output filelist can be generated using

python3 create_UL_campaign.py --mode preselection --era 2018 --task create_filelist --run all

Full Campaign

Setup

For the full campaign, two CMSSW versions are needed. They are setup using

python3 create_UL_campaign.py --mode full --era 2018 --task setup_cmssw --final-state $FINALSTATENAME

The possible Final State names are:

FINALSTATENAME = ["MuTau", "ElTau", "ElMu", "TauTau", "MuEmb", "ElEmb"]

After this setup, the two tarballs containing the CMSSW code are generated and uploaded to the grid storage using

python3 create_UL_campaign.py --mode full --era 2018 --task upload_tarballs --final-state $FINALSTATENAME

Job setup is done using

python3 create_UL_campaign.py --mode full --era 2018 --task setup_jobs --final-state $FINALSTATENAME --run all --workdir /path/to/workdir

for all runs of an era or by specified the name of the run instead of using all.

Production

The Production of a full campaign can be started using

python3 create_UL_campaign.py --mode full --era 2018 --task run_production --run all --final-state $FINALSTATENAME

This will automatically start the fitting grid control tasks.

Output Collection

Collecting the output files into a file list is done using

python3 create_UL_campaign.py --mode full --era 2018 --run all --workdir  /path/to/workdir --task create-filelist --final-state $FINALSTATENAME

For this step, it is nessessary to use a different environment, the script will say if a different environment is needed.

Publish Dataset

TODO

older campaigns

for older campaign, use the rereco branch