This package houses the configuration files used to run the art executables in the TDAQ system. The package also includes a python script used to generate the configuration files for the TDAQ art jobs starting from a single JSON file.
The main area houses a series of files used to configure the Online reconstruction and the trigger sequences:
trigProducers.fcl: module instances of the producer modules used in the Online reconstructiontrigSequences.fcl: sequences used in the trigger paths and Online reconstructiontrigFilters.fcl: module instances of the filter modules used in the Online reconstructiontrigServices.fcl: place holder for teh services used in the Online recotrigDigiInputEpilog.fcl: used only for Offline tests where the inout data are in a Digi format (not Fragments)
This directory contains the Menu configuration files that organize in a single JSON file the configuration of the several components needed to run the art-based steps of the TDAQ state machine: Online reconstruction, trigger selection, dataloggers, DQM
This directory houses the following scripts:
generateMenuFromJSON.py: it creates the set offclfiles necessary to configure using thedata/physMenu.jsonas input:- the trigger sequences that use the tracker+calorimeter data:
trigMenuPSConfig.fcl,trigMenu.fcl - the trigger sequences that use the CRV (+possibly the Trk+Cal) data:
aggMenuPSConfig.fcl,aggMenu.fcl - the datalogger:
trigLoggerConfig.fcl,trigLoggerMenu.fcl - the lumiLogger:
trigLumiLoggerConfig.fcl,trigLumiLoggerMenu.fclIt is also possible to specify only a specific event-mode in the trigger menu by using the option-evtMode; the possible options are:all,OnSpillandOffSpill. It is also possible to process a customized trigger-menu json file by using the option-mf.
- the trigger sequences that use the tracker+calorimeter data:
An example of command line is the following:
python mu2e_trig_config/python/generateMenuFromJSON.py -mf mu2e_trig_config/data/physMenu.json -o gen -evtMode all