Restore CSC DQM monitor outside Muon POG sequence

[fabiocos]

IN CMSSW_14_0_18 the PR #46298, backport of #46293, has been integrated with the purpose of alleviating the memory issues faced by Tier0 in express harvesting workflows. It turns out that his updated has removed also CSC DQM monitoring, as reported in https://cms-talk.web.cern.ch/t/missing-folder-in-streamexpress-run2024i-express-v2-dqmio/52718/5 , since this sequenced has been not appropriately embedded in the Muon POG sequence, see https://github.com/cms-sw/cmssw/pull/46298/files#r1796815700 .

@rseidita is aware of the issue, and agreed to relocate the involuntarily removed sequence so as to restore the functionality. This issue is meant to record the problem and track its solution. As soon as the fix is available and backported to the CMSSW_14_0_X branch, a patch will be needed to restore the missing functionality in the Tier0.

[cmsbuild]

cms-bot internal usage

[cmsbuild]

A new Issue was created by @fabiocos.

@Dr15Jones, @antoniovilela, @makortel, @mandrenguyen, @rappoccio, @sextonkennedy, @smuzaffar can you please review it and eventually sign/assign? Thanks.

cms-bot commands are listed here

[fabiocos]

@ptcox FYI

[fabiocos]

@ptcox FYI

[makortel]

assign dqm

[cmsbuild]

New categories assigned: dqm

@antoniovagnerini,@nothingface0,@rvenditti,@syuvivida,@tjavaid you have been requested to review this Pull request/Issue and eventually sign? Thanks

[fabiocos]

As an interim solution, pending a better organization of the muon DQM sequences to separate cscMonitor (and whatever else really needed) from the what can be really dropped without harm, I have opened #46364 , #46366 and #46365 to restore the previous situation as a clean basis, and allow the production of a quick patch release for Tier0 in 14_0_X @mandrenguyen @antoniovilela . This will unavoidably bring back the memory issues in T0 express harvesting, that were the reason for the change, but presumably to a smaller extent than adding directly the @muon sequence to the T0 configuration, that would activate additional folders and code filling them. The addition of the simple cscMonitor module to the T0 configuration is not possible, as it fails in the HARVESTING step configuration build (as observed independently while using cmsDriver.py within CMSSW).

[ptcox]

Hi everyone. I tried to look into this unwanted coupling with Muon POG DQM and would like to suggest the following:

1. In DQM/​CSCMonitorModule/​python/​csc_dqm_offlineclient_collisions_cff.py add cscMonitor just as is done for the TnPEfficiency sequence:
   cscOfflineCollisionsClients = cms.Sequence(dqmCSCOfflineClient+cscTnPEfficiencyClient) ->
   cscOfflineCollisionsClients = cms.Sequence(dqmCSCOfflineClient+cscMonitor+cscTnPEfficiencyClient)
2. Remove the occurrences of cscMonitor in DQMOffline/Muon:
   DQMOffline/​Muon/​python/​muonCosmicMonitors_cff.py - 3 occurrences
   DQMOffline/Muon/python/muonMonitors_cff.py - 1 occurrence
   DQMOffline/Muon/test/muonMonitors_cff.py - 1 occurrence
   There is no case in which cscMonitor should occur alone, without the rest of CSC offline DQM, so once it's added to cscOfflineCollisionsClients all should remain as it should as far as CSC DQM is concerned.
   I think this splits cscMonitor out of Muon POG DQM.
   Does anyone see anything wrong with this?

[rseidita]

Hi Tim, thanks a look for taking the time to look into this. It looks good to me; I will implement your suggestion, test it, and let you know

[ptcox]

Thanks for looking it. I'm now concerned I don't understand how things really work though - I am getting confused between cscTnPEfficiencyClient and cscTnPEfficiencyMonitor when trying to make the analogy with the TnPEfficiency.

[ptcox]

Yes, I am wrong: one needs a DQMEDHarvester to add to cscOfflineCollisionsClients. I don't know how CSCOfflineMonitor histograms get harvested.