Features needed for CMSSW_10_1_1

[fabiocos]

The validation of CMSSW_10_1_0 is ongoing, we already know that we need to back-port some fixes and some last minute planned additions. The target is a CMSSW_10_1_1 containing features already deployed and to be validated in 10_2_0_pre1 (the build and validation of these two releases will have to move unfortunately in parallel).

(Updated on April 7 23:00)

Here is the list of the PRs already backported from 10_2_X as fixes:

#22822 from cms-tsg-storm: Fix l1 emulation in HLT scripts (101X)
#22823 from folguera: BugFix for L3MuonCandidateProducerFromMuons
#22819 from bsunanda: Run2-alca126 Add eta dependent threshold for ECAL
#22818 from mmusich: [10.1.X] fix segmentation fault in SiStripBadComponentInfo (backport)
#22792 from thomreis: Fix L1T muon copy comparison DQM harvesting path - 101x
#22791 from mmusich: [10.1.X] remove useless triplication of AAG data in calibTree
#22800 from hroskes: fix preexisting tracker alignment validation

This issue is meant to collect the other features that are really necessary for the operations at the beginning of data taking and for the completion of the PF calibration.

The list that I am aware of at present is:

• Fix the product name for the PCL BeamSpot HP workflow #22827 Fix the product name for the PCL BeamSpot HP workflow (already in master)
• Steering scripts for O2Os #22831 O2O scripts (TO BE SIGNED, already in master, on hold pending clarification by @ggovi)
• Fix for the popcon procedure #22832 Fix for the popcon procedure (TO BE SIGNED, already in master)
• GT in autoCond for 10_1_1, combination of PR 22834 and 22893 #22897 GT in autoCond for 10_1_1, combination of PR 22834 and 22893
• DQM: SiPixelPhase1- Added Big Pixel Charge Plots [Backport for 10_1_X of #22324 ] #22844 DQM: SiPixelPhase1- Added Big Pixel Charge Plots [Backport for 10_1_X of DQM: SiPixelPhase1- Added Big Pixel Charge Plots #22324 ] (TO BE SIGNED, already in master)
• CTPPS: updated DQM for 2018 data taking (10_1_X) #22858 CTPPS: updated DQM for 2018 data taking (10_1_X) (TO BE SIGNED)
• ctpps pixel dqm update - backport #22866 ctpps dqm update - backport
• Do not throw exception for fitting error for CMSSW_10_1 #22860 Do not throw exception for fitting error for CMSSW_10_1
• Run2-hcx173 Backport #22835 #22857 (sunanda alcareco fix)
• SiPixelQuality backport for CMSSW 10_1_X: bugfix and memory cost mitigation #22854 (SiPixelQuality memory cost mitigation)
• Backport Changed CorrPCCProducer from EDProducer to DQMEDAnalyzer for LumiPOG #22870 (for Lumi PCC PCL workflow, TO BE SIGNED)
• Fixing bug in default values for EgammaHLTGsfTrackVarProducer : 10_1_X #22864 (bug fix in EgammaHLTGsfTrackVarProducer - backport)
• Migration of HLT to 1010 template (101X) #22824 (backport for HLT menu migration to 10_1 template - already in 10_2_X)
• Remove comparisons from GlobalAlgBlk equal operator - 101x #22794 Remove comparisons from GlobalAlgBlk equal operator - 101x
• L1T DQM Ignore BX range mismatch in OMTF to uGMT comparison - 101x #22793 L1T DQM Ignore BX range mismatch in OMTF to uGMT comparison - 101x
• L1T DQM FEDIntegrity update to stage2 - 101x #22795 L1T DQM FEDIntegrity update to stage2 - 101x

Please add further motivated requests to this issue

[cmsbuild]

A new Issue was created by @fabiocos Fabio Cossutti.

@davidlange6, @Dr15Jones, @smuzaffar, @fabiocos can you please review it and eventually sign/assign? Thanks.

cms-bot commands are listed here

[Dr15Jones]

@fabiocos #22860 is needed to avoid the exception seen in AlcaBeamMonitor

[fabiocos]

@popovvp @fabferro I understand that, besides #22858, for CTPPS we need also the backport of #22678 could you please provide it ?

[fabferro]

@fabiocos: yes we are going to open the backport of #22678 asap.

[lpernie]

From Alca side we need:
#22834 (2018 GT change)
#22857 (sunanda alcareco fix)
#22854 (SiPixelQuality memory cost mitigation)
#22858 and #22866 (ctpps fixes)
#22831 , #22832 (fix in O2O)
@ggovi , @arunhep you can cross-check this list.

[fabferro]

I've opened PR #22866 as a backport of PR #22678
@fabiocos @popovvp

[arunhep]

adding to what @lpernie already mentioned :
#22870 (for Lumi PCC PCL workflow)
and we are expecting a PR tomorrow which is needed to deploy new LHCInfo O2O
we will mention it here once it will be opened.

[Martin-Grunewald]

For TSG we request:

10_2_X/10_2_0_pre1
#22833 (2018 Conditions with new L1T menu)
#22863 (bug fix in EgammaHLTGsfTrackVarProducer)

10_1_X/10_1_1
#22834 (2018 Conditions with new L1T menu - backport)
#22864 (bug fix in EgammaHLTGsfTrackVarProducer - backport)
#22824 (backport for HLT menu migration to 10_1 template - already in 10_2_X)

[deguio]

hi @fabiocos
it would be good to have the following PRs merged as well.
#22848 (10_2_X)
#22849 (10_1_X)

which are a minor fix for the TP emulator for HCAL
thanks.

@christopheralanwest @abdoulline @mariadalfonso

[thomreis]

Hi @fabiocos
For L1T we would like to have these in 10_1_1 if possible.
#22794
#22793
#22795
#22880
@rekovic please take a look as well.

[fabiocos]

@deguio as we discussed at the operation meeting, #22849 can come at a later stage without rushing (not yet approved in master), so I will remove it from the list

[fabiocos]

@arunhep @lpernie Huston we have a problem...

After the integration in 10_2_X of the GT PR #22833 (whose backport #22834 should be one of the key residual ingredients for 10_1_1) we are getting several failures in tonight's IB, see for instance

https://cmssdt.cern.ch/SDT/cgi-bin/buildlogs/slc6_amd64_gcc630/CMSSW_10_2_X_2018-04-06-2300/pyRelValMatrixLogs/run/136.77_RunDoubleEG2016H+RunDoubleEG2016H+HLTDR2_2016+RECODR2_2016reHLT_skimDoubleEG_Prompt+HARVESTDR2/step3_RunDoubleEG2016H+RunDoubleEG2016H+HLTDR2_2016+RECODR2_2016reHLT_skimDoubleEG_Prompt+HARVESTDR2.log

They are reproducible, do not appear if only the GT PR is not used (In any case the other additions tonight look unrelated to this), and have the pattern

%MSG
----- Begin Fatal Exception 07-Apr-2018 07:13:54 CEST-----------------------
An exception of category 'Configuration' occurred while
[0] Processing Event run: 283877 lumi: 3 event: 3979066 stream: 2
[1] Running path 'pathALCARECOEcalESAlign'
[2] Calling method for module HLTHighLevel/'ALCARECOEcalESAlignHLT'
Exception Message:
HLTHighLevel [instance: ALCARECOEcalESAlignHLT - path: pathALCARECOEcalESAlign]: No triggerList with key EcalESAlign in AlCaRecoTriggerBitsRcd
----- End Fatal Exception -------------------------------------------------

A test on some of the failing workflows is running in #22834 to cross check that this appears also in the 10_1_X branch.

Any clue of what is wrong/missing? Needless to say that this is a show-stopper for 10_2_0_pre1 and 10_1_1 if not solved.

[arunhep]

@fabiocos i know the problem. i will open a PR soon.

[arunhep]

These PRs :
master : #22892
10.1.X : #22893
should fix the issue.

[fabiocos]

@arunhep #22897 combines #22834 and the fix #22893 (which should be rebased)

[fabozzi]

Hi, for PdmV it would be nice to have the updated GENSIM for relvals:
#22894 (master)
#22895 (101X)

[fabiocos]

@fabozzi the new datasets look all at CERN, so I expect we should have no failures due to missing inputs

[fabiocos]

At this point I will freeze the present situation, more additions may come in a following patch/release