Before changing the prescale column, remember to talk to the Shift Leader.

Prescale Column Choice instructions: PLEASE REFRESH AND READ AT START OF SHIFT. ALSO, LOOK AT THE SECTION FOR EXPLICIT INSTRUCTIONS - DO NOT GO BLINDLY BY TRIGGER COLUMN NAMES.

This twiki is maintained by FOG and all updates to this twiki should be agreed with FOG/the HLT DOC beforehand. The HLT DOC number is 165575 (or 66278).

note1: The rate limits are all for when T0 transfer is ON. If T0 transfer is OFF, then a special test is ongoing where the experts know they are exceeding agreed upon physics/express limits.

note2: If anything is unclear, please let the HLT DOC know so they can clarify it

last updated: 2nd June 2022

Summary

Name CMS Run Mode L1/HLT Mode Description Recommended Columns Notes
Cosmics cosmics cosmics2022 Cosmics menu, with core consisting of muon triggers Cosmics Standard recommended column.
HighRandom Commissioning: HighRandom is the recommended (and currently default) prescale column when running with high (~ 80-100 kHz) randoms during commissioning. Important: if the columns are changed to anything but HighRandom, the random trigger rate must be reduced to standard values (400 Hz). We do not have a duplicate of each column for the high randoms configuration.
Circulating circulating circulating2022 Circulating menu = cosmics + HLT_BptxOR Circulating Default column with prescale of 1103 on HLT_BptxOR
MediumBPTX, HighBPTX, VHighBPTX Special columns to handle higher number of circulating bunches (> 5, > 50, > 500).
Collisions (first 13.6 TeV) collisions collisions2022 HLT menu for 13.6 TeV collisions = minimal menu (ZeroBias + MinimumBias + AlCa) + HLT V1 (Run 3 physics menu) HalfZeroBiasReducedReco The default column is HalfZeroBiasReducedReco, which is also the recommended column for n(Bunches) = 2.
Emergency Commissioning: During fills with no actual collisions expected (eg. circulating beams during Adjust, Squeeze or Flat Top), the temporary recommendation is to stay in the Emergency column.
Collisions (8 bunches, 13.6 TeV) collisions collisions2022 HLT menu for 13.6 TeV collisions = minimal menu (ZeroBias + MinimumBias + AlCa) + HLT V1 (Run 3 physics menu) 8bReducedReco The default column is 8bReducedReco, which is also the recommended column for n(Bunches) = 8.
Emergency Commissioning: During fills with no actual collisions expected (eg. circulating beams during Adjust, Squeeze or Flat Top), the temporary recommendation is to stay in the Emergency column.

Important: Recommended does not mean the default column! It can be the case that it must be changed at the start of the run (as with the 900 GeV collisions menu).

Cosmics

The cosmics mode has prescale columns for a variety of scenarios. The most import thing is that the HLT rate of the Express stream should not exceed 250 Hz during cosmics and Physics should not exceed ~300Hz. Note: collision limits are different!

We expect that in normal operations the Physics and Express HLT rate to be ~200Hz with all muon systems. Note the only triggers in express for cosmics are

HLT_L1MASSUPT_0_0_10_v1, HLT_L1Mu0upt0_v1, HLT_L1Mu0upt20ip03_v1, HLT_L1Mu0upt20ip0_v1,HLT_L1Mu0upt20ip3_v1,HLT_L1QuadMu0OS_v1, HLT_L1SingleMuCosmics_v1, HLT_L1SingleMuOpen_DT_v2, HLT_L1SingleMuOpen_v2, HLT_L1SingleMuOpenupt5_v1, HLT_L1TripleMu0TriMass1to4_v1 and HLT_Random_v1

So any high express rate has to be either due to muons or randoms.

Column Description

column 0: Emergency

  • this column shuts off ECAL, HCAL and Pixel local reco at the HLT

column 1: Cosmics

  • the standard cosmics menu

column 2: NoGPU

  • the standard cosmics menu except that all paths which will offload to a GPU are disabled

column 3: HighRandom (default)

  • the standard cosmics menu expecting 70kHz of random triggers but can run up to 100kHz of random rate
  • random prescale is 10000
  • Important: if the columns are changed to anything but HighRandom, the random trigger rate must be reduced to standard values (400 Hz). We do not have a duplicate of each column for the high randoms configuration.

column 4: MediumRandom

  • the standard cosmics menu expecting 15kHz of random triggers but can run up to 30kHz of random rate
  • the random prescale is 2000

column 5: HighEG

  • the standard cosmics menu but with L1 EG seeds highly precaled
  • it also prescales ECAL local reco

column 6: HighJets

  • the standard cosmics menu but with L1 Jet seeds highly precaled
  • it also prescales HCAL local reco

column 7: HighCalo

  • the standard cosmics menu but with L1 EG and L1 Jet seeds highly precaled
  • it also prescales ECAL & HCAL local reco

Column Selection

We expect that cosmics running will be done with 70kHz to 100kHz of random triggers with the exact value being communicated by Run Coordination (the nominal value is currently 85kHz but again you should check with RC if in doubt as this is a parameter which changes frequently). Thus you should nominally start in column 3 which is now the default column. The expected HLT rates are ~200Hz for physics & express if all muon systems are in. If muon systems are out/off/onstandby, the rate will be lower.

If the express stream is >250Hz, there is a problem and you should check you are in the correct configuration as shown below! If you can not resolve this or are unsure what to do, please ring the HLT DOC (165575 or 66278) any time

Column selection check list (in absence of other special instructions):

  1. ask the DAQ shifter how many randoms they have selected
  2. this is typically 70kHz but may change via instructions from RC. Note if they answer 400Hz, ask if they are sure, currently we usually run the cosmics with high randoms.
  3. select the column based on the random rate
    1. for randoms ≤2000Hz, select column 1 "Cosmics"
    2. for randoms >2000Hz and ≤30kHz, select column 4: "MediumRandom"
    3. for randoms >30kHz, select column 3: "HighRandom" (default)

Other columns are for available for selection:

  • column 0 is only to be entered under the instructions of the HLT DOC
    • when entering this column, the DAQ shifter must adjust the random rate to 400Hz
  • column 2 is also only to be entered under the instructions of the HLT DOC
    • when entering this column, the DAQ shifter must adjust the random rate to 400Hz
  • columns 5,6,7 are for when when we have high rates (L1 and HLT rate > 300Hz) coming through on the L1 EG or L1 Jet seeds. This happens when there is a hot trigger tower in the detector in ECAL (EG + Jets) or HCAL (Jets only)
    • the shift crew can select this column on their own inititive and should do when significant rate is coming through on the EG and/or Jet seeds
    • when entering this column, the DAQ shifter must adjust the random rate to 400Hz
    • however the shift crew must follow the standard L1 instructions to resolve the issue (up to and including calling the L1 DOC) and leave this column as soon as it is resolved
    • this column is only intended to give experts time to resolve issues, it is not intended to be run in long term
    • during extended working hours (including weekends), if this column is remained in for more than ~30mins, the HLT DOC must be called
    • overnight, the HLT DOC is called at the discretion of the L1 DOC (as this is really a L1 issue not a HLT issue)

Circulating

The circulating menu is identical to the Cosmics menu, with the addition of the HLT_BptxOR path and three columns MediumBPTX, HighBPTX and VHighBPTX. In practice the circulating menu will give a lower rates than the cosmics menu, this is because the muon systems which provide almost all the cosmic menu triggers will usually be on standby when there is beam in the machine. If there is not beam in the machine and the muon systems are turned on, it is expected that you will be in the cosmics mode (please consult the SL instructions for further details). Thus this menu should only ever provide triggers from L1_EG seeds, L1_Jet seeds, Randoms, and HLT_BptxOR, of which HLT_BptxOR will drive the rate. Like the cosmics menu, only randoms and muons enter the express, and therefore given the muon systems will be off, the express rate exclusively comes from random triggers so any rate over ~7Hz means that the DAQ is sending more than the 400Hz randoms expected and it should be clarified why this is happening. If this random rate results in an express rate of more than 250Hz, T0 transfer is on, and this is an intentional operation, please make sure the HLT DOC is aware this is happening and is happy with it. If is unintentional, please ask the DAQ shifter to lower the randoms to the nominal value.

The circulating data does not take useful data and is only there to keep the system ticking over.

The rate of the physics stream will change with the LHC beam conditions, specifically how many bunches are in the machine. The column should be changed as described below to get the HLT physics stream in the range ~11-50Hz.

Column Description

column 0: Emergency

  • this column shuts off ECAL, HCAL and Pixel local reco at the HLT, as well as the jet paths.

column 1: Circulating (default and recommended)

  • the standard circulating menu

column 2: NoGPU

  • the standard circulating menu except that all paths which will offload to a GPU are disabled

column 3: HighRandom

  • the standard circulating menu expecting 70kHz of random triggers
  • random prescale is 10000

column 4: MediumRandom

  • the standard circulating menu expecting 15kHz of random triggers
  • the random prescale is 2000

column 5: HighEG

  • the standard circulating menu but with L1 EG seeds highly prescaled
  • it also prescales ECAL local reco

column 6: HighJets

  • the standard circulating menu but with L1 Jet seeds highly prescaled
  • it also prescales HCAL local reco

column 7: HighCalo

  • the standard circulating menu but with L1 EG and L1 Jet seeds highly prescaled
  • it also prescales ECAL & HCAL local reco

column 8: MediumBPTX

  • the standard circulating menu but with HLT_BptxOR prescaled by 11030 at L1

column 9: HighBPTX

  • the standard circulating menu but with HLT_BptxOR prescaled by 110300 at L1

column 10: VHighBPTX

  • the standard circulating menu but with HLT_BptxOR prescaled by 1103000 at L1

Column Selection

Column Selection check list (in absence of other special instructions):

  1. confirm that the DAQ shifter has specified 400Hz of randoms
  2. select one of columns 1, 8, 9 or 10 such that the rate of the HLT physics stream is between ~11 to ~50 Hz
    • each subsequent column increases the prescale of HLT_BptxOR by 10
    • the ~11 to 50Hz range is actually a translation of keeping HLT_BptxOR between 4 to 43Hz as there is 7Hz of random triggers also present
    • if moving up a column does not fix the physics rate, then assuming the randoms are correctly set at 400Hz, the L1 EG and L1 Jet triggers are giving the improper rate, please resolve this issue as per the standard L1 instructions
    • if the BptxOR rate allows it, it is permissible to move into columns 5,6,7 in this case while experts fix the problem

The other columns are for the cosmics menu and are not intended to be selected by the shift crew in circulating unless instructed to by the HLT DOC.

The express rate in HLT should always be ~7Hz as the muon systems are off. If this is not 7Hz, either the random rate is improperly set or the muon systems are sending triggers which should not be the case as they are expected to be off. If the random rate is higher than 400Hz randoms expected and it should be clarified why this is happening. If this random rate results in an express rate of more than 250Hz, T0 transfer is on, and this is an intentional operation, please make sure the HLT DOC is aware this is happening and is happy with it. If is unintentional, please ask the DAQ shifter to lower the randoms to the nominal value.

Collisions (pp, 13.6 TeV with more than 2 bunches)

This is describing the 13.6 TeV collisions menu. We will use this for all collision runs from July 1st onwards, even for those at 900 GeV.

  • The ZeroBias rate is 11.2 kHz per colliding bunch.

The menu primarily writes out ZeroBias (all colliding bunches). The full ZeroBias rate is split between 20 datasets. It also runs paths necessary for AlCa. We are also running the nominal HLT physics paths as a test, even though they may have a very low rate.

The correct choice of column will depend on where we are in commissioning the trigger system and the number of bunches in the machine. Therefore the HLT DOC will tell you the correct column, please contact them if you are unsure.

For the random rate, please check with Shift Leader what are the instructions from RC and HLT. The current recommandation for 8 colliding bunches is to set the random rate to 20kHz.

Note: For 4 or less colliding bunches, the rule of thumb is for the DAQ shifter to set the random trigger rate to 10 kHz/colliding bunch, in order to reduce the ReTri deadtime, which protects the tracker from regular triggers.

  • For 2 colliding bunches, the random trigger rate should be set to 20 kHz.
    • Therefore in this case, the L1 input rate equals 42.5 kHz, i.e. the sum of the ZeroBias rate (22.5 kHz, for 2 colliding bunches) and random trigger rate (20 kHz).

Side Note: During beam commissioning, and particularly during unstable collision, there may be possible high rates from the JetHT Primary Dataset. A good indicator of this happening is if the rate of HLT_AK8PFJetFwd15 goes above 2 kHz. If you observe this is is the case, go to one of the "ReducedReco" columns, where those paths are not run. Raise a warning about this in the Shifters+DOCs Mattermost room.

Column Description

Please confirm with the HLT DOC (165575 or 66278) which column to be in, as it will be changing during the commissioning period. The default column is column 3 8bReducedReco, and this is the recommended column for the stable fills with n(Bunches) = 8. This writes 1/9th of all Zero Bias data, and runs part of the HLT reconstruction (e.g. AlCa triggers). For stable beams with n(Bunches)=2, the recommended column is 1 (2bReducedReco). If n(Bunches) > 8 are expected (or occur during the run), please go to column 0 (Emergency), and call the HLT DOC.

During actual collisions, you have to adapt to the status of CMS commissioning activities; choose one of the columns below and, again, if needed confirm with the HLT DOC. A summary of each prescale column is given in the following format:

ZeroBias un/prescaled; AlCa paths enabled/disabled; physics paths enabled/disabled

  • The prescale values are given in the format (L1_PS*HLT_PS)
  • The expected rates can be calculated with the expected Zero/MinimumBias rates (above) and their corresponding prescale

column 0: Emergency

  • It is a column for standby use. It only writes out randoms and ECAL and HCAL calibration triggers, i.e. it effectively disables the CMS output.
    • Therefore the rate will be very small so as not to risk accidentally overloading T0
    • During actual collisions you definitely don't want to be in this column!
  • ZeroBias prescaled by (1103*100); AlCa paths disabled; physics paths disabled
  • The DST_Physics_v7 (PS=1) and HLT_Random_v3 (PS=1800) triggers and HLT_EcalCalibration and HLT_HcalCalibration are enabled

column 1: 2bReducedReco (recommended column for n(Bunches)=2)

  • Writes out half of the ZeroBias rate and AlCa paths.
  • ZeroBias prescaled by (1*2); AlCa paths enabled; physics paths disabled

column 2: 8bNoReco

  • Writes out 1/9th of the ZeroBias rate only.
  • ZeroBias prescaled by (3*3); AlCa paths disabled; physics paths disabled

column 3: 8bReducedReco (default column, recommended for n(Bunches)=8)

  • Writes out 1/9th of the ZeroBias rate and AlCa paths.
  • ZeroBias prescaled by (3*3); AlCa paths enabled; physics paths disabled

column 4: 8bFullReco

  • Writes out 1/9th of the ZeroBias rate, AlCa paths, and all standard physics paths.
  • ZeroBias prescaled by (3*3); AlCa paths enabled; physics paths enabled

Collisions (pp, first 13.6 TeV)

This is describing the 13.6 TeV collisions menu. We will use this for all collision runs from July 1st onwards, even for those at 900 GeV.

  • The ZeroBias rate is 11.2 kHz per colliding bunch.

The menu primarily writes out ZeroBias (all colliding bunches). The full ZeroBias rate is split between 20 datasets. It also runs paths necessary for AlCa. We are also running the nominal HLT physics paths as a test, even though they may have a very low rate.

The correct choice of column will depend on where we are in commissioning the trigger system and the number of bunches in the machine. Therefore the HLT DOC will tell you the correct column, please contact them if you are unsure.

Side Note: As a rule of thumb (up to 4 colliding bunches) the random trigger rate should be set (by the DAQ shifter) to 10 kHz/colliding bunch, in order to reduce the ReTri deadtime, which protects the tracker from regular triggers.

  • For 2 colliding bunches, the random trigger rate should be set to 20 kHz.
    • Therefore, the L1 input rate is the sum of the ZeroBias rate (22.5 kHz, for 2 colliding bunches) and random trigger rate (20 kHz) = 42.5 kHz.

Side Note 2: During beam commissioning, and particularly during unstable collision, there may be possible high rates from the JetHT Primary Dataset. A good indicator of this happening is if the rate of HLT_AK8PFJetFwd15 goes above 2 kHz. If you observe this is is the case, go to one of the "ReducedReco" columns, where those paths are not run. Raise a warning about this in the Shifters+DOCs Mattermost room.

Column Description

Please confirm with the HLT DOC (165575 or 66278) which column to be in, as it will be changing during the commissioning period. The default column is column 2 HalfZeroBiasReducedReco, and this is the recommended column for the stable fills with n(Bunches) = 2. This writes half of all Zero Bias data, and runs part of the HLT reconstruction (e.g. AlCa triggers). For stable beams with n(Bunches)=4, column 3 (QuarterZeroBiasReducedReco) can be used. If n(Bunches) > 4 are expected, please call the HLT DOC. The HLT menu and this Twiki will be updated soon for scenarios with n(Bunches)>>4.

During actual collisions, you have to adapt to the status of CMS commissioning activities; choose one of the columns below and, again, if needed confirm with the HLT DOC. A summary of each prescale column is given in the following format:

ZeroBias un/prescaled; AlCa paths enabled/disabled; physics paths enabled/disabled

  • The prescale values are given in the format (L1_PS*HLT_PS)
  • The expected rates can be calculated with the expected Zero/MinimumBias rates (above) and their corresponding prescale

column 0: Emergency

  • It is a column for standby use. It only writes out randoms and ECAL and HCAL calibration triggers, i.e. it effectively disables the CMS output.
    • Therefore the rate will be very small so as not to risk accidentally overloading T0
    • During actual collisions you definitely don't want to be in this column!
  • ZeroBias prescaled by (1103*100); AlCa paths disabled; physics paths disabled
  • The DST_Physics_v7 (PS=1) and HLT_Random_v3 (PS=1800) triggers and HLT_EcalCalibration and HLT_HcalCalibration are enabled

column 1: ZeroBiasReducedReco

  • Writes out the full ZeroBias rate and AlCa paths.
  • ZeroBias unprescaled; AlCa paths enabled; physics paths disabled

column 2: HalfZeroBiasReducedReco (default column, recommended for n(Bunches)=2)

  • Writes out half of the ZeroBias rate and AlCa paths.
  • ZeroBias prescaled by (1*2); AlCa paths enabled; physics paths disabled

column 3: QuarterZeroBiasReducedReco

  • Writes out one quarter of the ZeroBias rate and AlCa paths.
  • ZeroBias prescaled by (1*4); AlCa paths enabled; physics paths disabled

column 4: HighPrescaleZeroBiasReducedReco

  • Writes out 1/9th of the ZeroBias rate and AlCa paths.
  • ZeroBias prescaled by (3*3); AlCa paths enabled; physics paths disabled

column 5: ZeroBiasNoReco

  • Writes out the full ZeroBias rate only.
  • ZeroBias unprescaled; AlCa paths disabled; physics paths disabled

column 6: HalfZeroBiasNoReco

  • Writes out half of the ZeroBias rate only.
  • ZeroBias prescaled by (1*2); AlCa paths disabled; physics paths disabled

column 7: QuarterZeroBiasNoReco

  • Writes out one quarter of the ZeroBias rate only.
  • ZeroBias prescaled by (1*4); AlCa paths disabled; physics paths disabled

column 8: HighPrescaleZeroBiasNoReco

  • Writes out 1/9th of the ZeroBias rate only.
  • ZeroBias prescaled by (3*3); AlCa paths disabled; physics paths disabled

column 9: HalfZeroBiasFullReco

  • Writes out half of the ZeroBias rate, AlCa paths, and all standard physics paths.
  • ZeroBias prescaled by (1*2); AlCa paths enabled; physics paths enabled

column 10: L1BunchMaskNoReco

  • Output ZeroBias but instead of using the ZeroBias trigger, it uses L1_AlwaysTrue masked by the ALGOBX_MASK parameter which allows us to manually specify the bunches to trigger on.
  • Used in case of a problem with the BRIL system and for testing and should only be used under expert advice
  • Same as column 5.
  • ZeroBias unprescaled; AlCa paths disabled; physics paths disabled
    • Note: ZeroBias = HLT_ZeroBias seeded by L1_AlwaysTrue * ALGOBX_MASK

column 11: L1BunchMaskFullReco

  • Output ZeroBias but instead of using the ZeroBias trigger, it uses L1_AlwaysTrue masked by the ALGOBX_MASK parameter which allows us to manually specify the bunches to trigger on.
  • Used in case of a problem with the BRIL system and for testing and should only be used under expert advice
  • Same as column 10, but turns on all AlCa paths and standard physics paths.
  • ZeroBias unprescaled; AlCa paths enabled; physics paths enabled
    • Note: ZeroBias = HLT_ZeroBias seeded by L1_AlwaysTrue * ALGOBX_MASK

Collisions (pp, 900 GeV)

This is describing the 900 GeV collisions menu. Due to the extremely low pileup (PU ~ 0.05), the HLT can take high rates. We expect an HLT rate up to 45 kHz! This does not apply normally!

  • The ZeroBias rate is 11.2 kHz per colliding bunch. The expected MinimumBias (PU ~ 0.05) rates are of the order of 250 - 600 Hz per colliding bunch.

The menu primarily writes out ZeroBias (all colliding bunches) and MinimumBias (pass though of L1_MinimumBiasHF0_AND_BptxAND). The ZeroBias is split between 20 datasets while the MinimumBias is split between 10 in order not to surpass per dataset size limit. It also runs paths necessary for AlCa. We are also running the nominal HLT physics paths as a test, even though they have a very low rate.

The correct choice of column will depend on where we are in commissioning the trigger system and the number of bunches in the machine. Therefore the HLT DOC will tell you the correct column, please contact them if you are unsure.

Side Note: As a rule of thumb (up to 4 colliding bunches) the random trigger rate should be set (by the DAQ shifter) to 10 kHz/colliding bunch, in order to reduce the ReTri deadtime, which protects the tracker from regular triggers.

  • For 2 colliding bunches, the random trigger rate should be set to 20 kHz.
    • Therefore, the L1 input rate is the sum of the ZeroBias rate (22.5 kHz, for 2 colliding bunches) and random trigger rate (20 kHz) = 42.5 kHz.

Side Note 2: During beam commissioning, and particularly during unstable collision, there may be possible high rates from the JetHT Primary Dataset. A good indicator of this happening is if the rate of HLT_AK8PFJetFwd15 goes above 2 kHz. If you observe this is is the case, go to column 16 (ZeroBiasHalfMinBias), since it also prescales jet paths by a large factor. Raise a warning about this in the Shifters+DOCs Mattermost room.

Column Description

Please confirm with the HLT DOC (165575 or 66278) which column to be in, as it will be changing during the commissioning period. The recommended column is column 1 ZeroBiasMinBias for the stable fills with n(Bunches) ≤ 4. The default column is set to 0 Emergency, which disables output, and should be changed during actual collisions

A summary of each prescale column is given in the following format:

ZeroBias un/prescaled; MinimumBias un/prescaled or disabled; AlCa paths enabled/disabled; physics paths enabled/disabled

  • The prescale values are given in the format (L1_PS*HLT_PS)
  • The expected rates can be calculated with the expected Zero/MinimumBias rates (above) and their corresponding prescale

column 0: Emergency (default column)

  • It is a column for standby use. It only writes out randoms and ECAL and HCAL calibration triggers, i.e. it effectively disables the CMS output.
    • Therefore the rate will be very small so as not to risk accidentally overloading T0
  • It is now the default column as there is concern in other collision columns, the LHC could potentially give us a very high rate when they are commissioning with circulating coinciding bunches.
    • During actual collisions you definitely don't want to be in this column!
  • ZeroBias prescaled by (1103*100); MinimumBias disabled; AlCa paths disabled; physics paths disabled
    • Note: L1_MinimumBiasHF prescaled by 500 @ L1T
  • The DST_Physics_v7 (PS=1) and HLT_Random_v3 (PS=1800) triggers and HLT_EcalCalibration and HLT_HcalCalibration are enabled

column 1: ZeroBiasMinBias (recommended column for n(Bunches) ≤ 4)

  • Writes out all ZeroBias and MinimumBias datasets, while all physics paths are unprescaled (except of those seeded by L1_Random or L1_ZeroBias)
  • ZeroBias unprescaled; MinimumBias unprescaled; AlCa paths enabled; physics paths enabled

column 2: ZeroBias

  • Writes out the ZeroBias but not MinimumBias
  • ZeroBias unprescaled; MinimumBias disabled; AlCa paths enabled; physics paths enabled

column 3: MinBias (recommended column for 4 < n(Bunches) ≤ 8)

  • Writes out the MinimumBias but prescales the ZeroBias by a factor 2, when we can not record all events (expected when 4 < n(Bunches) ≤ 8)
  • ZeroBias prescaled by (1*2); MinimumBias unprescaled; AlCa paths enabled; physics paths enabled

column 4: MinBiasMedium (recommended column for n(Bunches) > 10)

  • Writes out the MinimumBias but prescales the ZeroBias by a factor 6, when we can not record all events (expected when n(Bunches) > 10)
  • ZeroBias prescaled by (3*2); MinimumBias unprescaled; AlCa paths enabled; physics paths enabled
    • Note: PixelClusters MinBias paths seeded by 'OR' of unprescaled L1_MinimumBiasHF0_AND_BptxAND and prescaled ZeroBias

column 5: MinBiasHigh

  • Writes out the MinimumBias but prescales the ZeroBias by a factor 10, when we can not record all events (expected when n(Bunches) > 24)
  • ZeroBias prescaled by (5*2); MinimumBias unprescaled; AlCa paths enabled; physics paths enabled
    • Note: PixelClusters MinBias paths seeded by 'OR' of unprescaled L1_MinimumBiasHF0_AND_BptxAND and prescaled ZeroBias

column 6: MinBiasVeryHigh

  • When we can no longer take all L1_MinimumBiasHF0_AND_BptxAND events at HLT
    • HLT pass through for L1_MinimumBiasHF0_AND_BptxAND prescaled by factor 5
    • relies on HLT_PixelClusters_WP2 to collect data
  • ZeroBias prescaled by (11*10); MinimumBias prescaled by 5; AlCa paths enabled; physics paths enabled
    • Note: PixelClusters MinBias paths seeded by 'OR' of prescaled L1_MinimumBiasHF0_AND_BptxAND and prescaled ZeroBias

column 7: ZeroBiasMinBiasReducedReco

  • Same as column 1 but disables physics HLT paths
    • equivalent to the minimal menu with only ZeroBias + MinimumBias + AlCa
  • ZeroBias unprescaled; MinimumBias unprescaled; AlCa paths enabled; physics paths disabled

column 8: ZeroBiasReducedReco

  • Same as column 2 but disables physics HLT paths
  • ZeroBias unprescaled; MinimumBias disabled; AlCa paths enabled; physics paths disabled

column 9: MinBiasReducedReco

  • Same as column 3 but disables physics HLT paths
  • ZeroBias prescaled by (1*2); MinimumBias unprescaled; AlCa paths enabled; physics paths disabled

column 10: ZeroBiasMinBiasNoReco

  • Same as column 1 but disables all reconstruction
  • ZeroBias unprescaled; MinimumBias unprescaled; AlCa paths disabled; physics paths disabled

column 11: ZeroBiasNoReco

  • Same as column 2 but disables all reconstruction
  • ZeroBias unprescaled; MinimumBias disabled; AlCa paths disabled; physics paths disabled

column 12: MinBiasNoReco

  • Same as column 3 but disables all reconstruction
  • ZeroBias prescaled by (1*2); MinimumBias unprescaled; AlCa paths disabled; physics paths disabled

column 13: L1BunchMaskNoReco

  • Same as column 14 but disables all reconstruction
  • ZeroBias unprescaled; MinimumBias unprescaled; AlCa paths disabled; physics paths disabled
    • Note: ZeroBias = HLT_ZeroBias seeded by L1_AlwaysTrue * ALGOBX_MASK

column 14: L1BunchMask

  • Output ZeroBias but instead of using the ZeroBias trigger, it uses L1_AlwaysTrue masked by the ALGOBX_MASK parameter which allows us to manually specify the bunches to trigger on
  • Used in case of a problem with the BRIL system and for testing and should only be used under expert advice
  • MinimumBias is disabled although HLT_PixelClusters_WP2 paths are still active
  • Disables all L1 seeds except L1_AlwaysTrue and L1_CDC_SingleMu_3_er1p2_TOP120_DPHI2p618_3p142
  • ZeroBias unprescaled; MinimumBias unprescaled; AlCa paths enabled; physics paths enabled
    • Note: ZeroBias = HLT_ZeroBias seeded by L1_AlwaysTrue * ALGOBX_MASK

column 15: ZeroBiasMinBiasHighReco

  • Same as column 1 but with scouting path enabled (DST_Run3_PFScoutingPixelTracking) enabled
  • ZeroBias unprescaled; MinimumBias unprescaled; AlCa paths enabled; physics paths enabled

column 16: ZeroBiasHalfMinBias

  • Same as column 1 but ZeroBias is prescaled by 2 and MinBias by 4 at HLT level. This means that it will write out half of ZeroBias and a quarter of MinBias data.
  • ZeroBias prescaled by (1*2); MinimumBias prescaled by (1*4); AlCa paths enabled; physics paths enabled
  • Jet paths prescaled by a large factor

column 17: ZeroBiasQuarterMinBias

  • Same as column 1 but ZeroBias is prescaled by 4 and MinBias by 8 at HLT level. This means that it will write out a quarter of ZeroBias and an eighth of MinBias data.
  • ZeroBias prescaled by (1*4); MinimumBias prescaled by (1*8); AlCa paths enabled; physics paths enabled
  • Jet paths prescaled by a large factor

High Rate Tests for Individual Subsystems

From time to time subsystems will be involved in high L1 rate tests (>70kHz) with just that subsystem in the run. The menu used can be either cosmics or circulating. The test must be done in column 7 with Tier0 transfer OFF. High physics and express rates will be observed during this but with Tier0 transfer OFF, this is not a problem. If a system wishes to do a high rate test with only them included in the DAQ and have tier 0 transfer on, this must be announced to the HLT DOC ahead of time who can come up with a solution.

The reason for this is that DAQ3 dynamically sizes the HLT farm based on included subsystems. For example when just ECAL or HCAL is included, the farm size is too small to run ECAL or HCAL local reconstruction at a high rate. Thus the HLT will give backpressure and the DAQ will warn that HLT CPU is saturating. Column 7 prescales the ECAL local reco by 100 and the HCAL local reco by 1000 reducing the required CPU to manageable levels.

Column Selection:

Column Selection check list (in absense of other special instructions):

  1. confirm with DAQ shifter that Tier0 transfer is OFF
  2. select column 7 HighCalo

How to Change the Prescale Column

Change of prescale column index during a run

  1. Go to the uGT, commands panel ControlPanels -> uGT Prescales
  2. Click on "Refresh Table". Check the Current Prescale index at the top of the panel.
  3. Click on the column to be chosen.
  4. If the trigger is reconfigured, this needs to be repeated after the reconfiguration.

Edit | Attach | Watch | Print version | History: r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r1 - 2022-07-06 - MarinoMissiroli
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Sandbox All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback