Skip to content

Reimplement Run1+Run2 functionality for upgrade

This issue lists the Brunel functionality that was being tested by the Brunel Run1 + Run2 tests, which are being removed from Brunel/master (see !777 (merged)). Some or all of these tests should be replaced when equivalent functionality is implemented in the Run 3 reconstruction, if still relevant. This continues the work started in LHCBPS-1801

  • Simulation = True/False (switches between using SIMCOND and LHCBCOND+ONLINE) . Currently all upgrade tests use conditions from SIMCOND/Upgrade. It will be mandatory to check that all the necessary conditions exist for the reconstruction to work when Simulation=False

  • Input Data Type - can be DIGI, XDIGI, MDF (real data), DST, XDST . Currently only DIGI and XDIGI are tested. MDF tests ability to reconstruct real data, DST and XDST ability to start reconstruction from a DST. Any future file format that contains RawEvent should be tested.

  • Output Data Type - can be MDF, RDST, DST, XDST . Currently only DST and XDST are tested. Any format needed for Run3 offline workflows should be tested

  • Production Mode = True/False - tests configuration needed for running in Production (minimal output to log file, production of XML summary). . Currently only False tested

  • Histograms - various levels of monitoring possible, all required functionality to be tested (e.g. Online monitoring, online and offline DQ, simulation validation (“Checking”))

  • SpecialData

    • “FieldOff” - checks reconstruction without magnetic field. (requires field off data)
    • “Cosimcs” - special set up for tracking with cosmics. Not implemented for upgrade tracking detectors
    • “pGun” - reconstruct particle gun montecarlo
    • “pA” - special configuration for IFT reconstruction

  • Special configurations

    • VetoHltErrorEvents - check that Brunel can process Hlt passthrough events (i.e. without HltDecReports)
    • Hlt{1,2}FilterCode - allows to apply a filter to events to be processed by Brunel
    • SkipTracking - reconstruct only PID detectors, starting from a DST containing tracks

  • Special tests

    • sim08-invfield-down.qmt, sim08-invfield-up.qmt - check reconstruction is identical when reconstructing the same dataset with wrong (equal&opposite) magnetic field
    • initial-time.qmt - reruns an existing test (2017magup) with a EventClockSvc().InitialTime outside of data taking, to check that database IOV handling is working correctly (reconstruction should be identical. Requires LHCBCOND+ONLINE, i.e. Simulation=False)
    • MessageSvc().countInactive=True - looks for any unprotected DEBUG and VERBOSE messages
    • repro2012magdown - reruns on output file of 2012magdown test, to check that reprocessing gives identical reconstruction to first pass
    • readdstfsr - Verifies that FSRs are produced on a reconstructed DST. Uses 2017-magup as input
    • test-genfsr - Check that Gen FSR (dumped to xml file) is correct
Edited by Vava Gligorov