Skip to content

Draft: WIP: PEB for AFP and Jets - v2

Marco Montella requested to merge mmontell/athena:PEB-Dev into master

This MR is the newer version of !49332 (closed) (now tracking my fork instead of Caterina's) and contains changes by Caterina from two previous separate MRs (now closed) to test new PEB (+TLA) trigger combinations:

The individual chains work when we run one at a time with commands like:

athenaHLT.py -l DEBUG -c "setMenu='Dev_pp_run3_v1';doL1Sim=True;rewriteLVL1=True; selectChains=['HLT_j20_JetPEBPhysicsTLA_L1J100'];" -o output --imf --perfmon --threads=1 --concurrent-events=1 --nprocs=1 --number-of-events=100 --file=/cvmfs/atlas-nightlies.cern.ch/repo/data/data-art/TrigP1Test/data18_13TeV.00360026.physics_EnhancedBias.merge.RAW._lb0151._SFO-1._0001.1 TriggerJobOpts/runHLT_standalone.py | tee out.txt

Likewise a test run with a partial TLA and PEB menu works fine:

athenaHLT.py -l WARNING -c "selectChains=['HLT_noalg_AFPPEB_L1EM22VHI','HLT_j20_PhysicsTLA_L1J100', 'HLT_j20_PhysicsTLA_L1J50_DETA20-J50J', 'HLT_j20_PhysicsTLA_L1HT190-J15s5pETA21', 'HLT_j20_PhysicsT\ LA_L1jJ100', 'HLT_j20_PhysicsTLA_L1jJ50_DETA20-jJ50J', 'HLT_j20_PhysicsTLA_L1HT190-jJ15s5pETA21', 'HLT_j20_JetPEBPhysicsTLA_L1J100'];doL1Sim=True;setMenu='Dev_pp_run3_v1';" -o output --imf --perfmon \ --threads=1 --concurrent-events=1 --nprocs=1 --number-of-events=10 --dump-config-reload --file=/cvmfs/atlas-nightlies.cern.ch/repo/data/data-art/TrigP1Test/data18_13TeV.00360026.physics_EnhancedBias.\ merge.RAW._lb0151._SFO-1._0001.1 TriggerJobOpts/runHLT_standalone.py

However there is currently a problem with ComboMaker that we don't understand, when running the entire menu. The error can be reproduced as follows:

  • release: 2022-01-09T2101
  • command: athenaHLT.py -l DEBUG -c "setMenu='Dev_pp_run3_v1';doL1Sim=True;rewriteLVL1=True; " -o output --imf --perfmon --threads=1 --concurrent-events=1 --nprocs=1 --number-of-events=100 --file=/cvmfs/atlas-nightlies.cern.ch/repo/data/data-art/TrigP1Test/data18_13TeV.00360026.physics_EnhancedBias.merge.RAW._lb0151._SFO-1._0001.1 TriggerJobOpts/runHLT_standalone.py | tee out.txt
  • error snippet in [1]
  • full output: /afs/cern.ch/user/d/doglioni/public/PEB/outPEBTests.txt

This also shows up in the ART tests when we run them ourselves. This error is discussed in ATR-24779

@doglioni and @mmontell

[1]

2021-12-23 20:55:38,235  Py:TriggerMenuMT:HLTCFConfig         DEBUG Adding chain HLT_j20_JetPEBPhysicsTLA_L1J100 to input HLTNav_ComboHypo_MainStep_jet_a4_tc
_em_subjesIS_0 of FStep18_merged1_TLAStep_MainStep_jet_a4_tc_em_subjesIS
2021-12-23 20:55:38,235  Py:TriggerMenuMT:HLTCFConfig         DEBUG Now Filter has chains: ['HLT_j20_PhysicsTLA_L1J100', 'HLT_j20_PhysicsTLA_L1J50_DETA20-J50
J', 'HLT_j20_PhysicsTLA_L1HT190-J15s5pETA21', 'HLT_j20_PhysicsTLA_L1jJ100', 'HLT_j20_PhysicsTLA_L1jJ50_DETA20-jJ50J', 'HLT_j20_PhysicsTLA_L1HT190-jJ15s5pETA2
1', 'HLT_j20_JetPEBPhysicsTLA_L1J100']
2021-12-23 20:55:38,235  Py:TriggerMenuMT:HLTCFConfig         DEBUG Now Filter has chains/input: [['HLT_j20_PhysicsTLA_L1J100', 'HLT_j20_PhysicsTLA_L1J50_DET
A20-J50J', 'HLT_j20_PhysicsTLA_L1HT190-J15s5pETA21', 'HLT_j20_PhysicsTLA_L1jJ100', 'HLT_j20_PhysicsTLA_L1jJ50_DETA20-jJ50J', 'HLT_j20_PhysicsTLA_L1HT190-jJ15
s5pETA21', 'HLT_j20_JetPEBPhysicsTLA_L1J100']]
2021-12-23 20:55:38,236  Py:TriggerMenuMT:MenuComponents      ERROR ComboMaker for Alg:ComboHypo_Step2_merged1_TLAStep_MainStep_jet_a4_tc_em_subjesIS with ad
dChain for:HLT_j20_JetPEBPhysicsTLA_L1J100 Chain multiplicity:(1,) Per leg input collection index:(0, 0).
2021-12-23 20:55:38,236  Py:TriggerMenuMT:MenuComponents      ERROR The size of the multiplicies vector must be the same size as the per leg input collection
 vector.  
2021-12-23 20:55:38,236  Py:TriggerMenuMT:MenuComponents      ERROR The ComboHypo needs to know which input DecisionContainers contain the DecisionObjects to
 be used for each leg.
2021-12-23 20:55:38,236  Py:TriggerMenuMT:MenuComponents      ERROR Check why ComboMaker.addInput(...) was not called exactly once per leg.
Shortened traceback (most recent user call last):
Shortened traceback (most recent user call last):
  File "/cvmfs/atlas-nightlies.cern.ch/repo/sw/master_Athena_x86_64-centos7-gcc11-opt/2021-12-22T2101/Athena/22.0.51/InstallArea/x86_64-centos7-gcc11-opt/job
Options/TrigPSC/TrigPSCPythonSetup.py", line 108, in <module>
    include( "%s" % PscConfig.optmap['JOBOPTIONSPATH'] )
  File "/cvmfs/atlas-nightlies.cern.ch/repo/sw/master_Athena_x86_64-centos7-gcc11-opt/2021-12-22T2101/Athena/22.0.51/InstallArea/x86_64-centos7-gcc11-opt/job
Options/TriggerJobOpts/runHLT_standalone.py", line 519, in <module>
    menu.generateMT()
  File "/scratch/doglioni/build_TLAPEB/build/x86_64-centos7-gcc11-opt/python/TriggerMenuMT/HLTMenuConfig/Menu/GenerateMenuMT.py", line 491, in generateMT
    makeHLTTree(newJO=False, triggerConfigHLT = TriggerConfigHLT)
  File "/scratch/doglioni/build_TLAPEB/build/x86_64-centos7-gcc11-opt/python/TriggerMenuMT/HLTMenuConfig/Menu/HLTCFConfig.py", line 152, in makeHLTTree
    finalDecisions = decisionTreeFromChains(steps, triggerConfigHLT.configsList(), triggerConfigHLT.dictsList(), newJO)
  File "/scratch/doglioni/build_TLAPEB/build/x86_64-centos7-gcc11-opt/python/TriggerMenuMT/HLTMenuConfig/Menu/HLTCFConfig.py", line 346, in decisionTreeFromC
hains
    (finalDecisions, CFseq_list) = createDataFlow(chains, allDicts)
  File "/scratch/doglioni/build_TLAPEB/build/x86_64-centos7-gcc11-opt/python/TriggerMenuMT/HLTMenuConfig/Menu/HLTCFConfig.py", line 449, in createDataFlow
    chainStep.combo.addChain( [d for d in allDicts if d['chainName'] == chain.name ][0])
  File "/scratch/doglioni/build_TLAPEB/build/x86_64-centos7-gcc11-opt/python/TriggerMenuMT/HLTMenuConfig/Menu/MenuComponents.py", line 336, in addChain
    raise Exception("[createDataFlow] Error in ComboMaker.addChain. Cannot proceed.")
 Exception: [createDataFlow] Error in ComboMaker.addChain. Cannot proceed.
Edited by Jakub Kremer

Merge request reports