Drop of D49-D85 Phase-2 geometry

[srimanob]

Following the HGCAL proposal in https://docs.google.com/document/d/1fCIpJJQLw0vGCKyR9LRmFmWVcAQZOlHk5jK-ESQLJnQ/edit?usp=sharing, they propose to preserve only V16 and V17 of HGCAL (== C17-C19).

So the affected geometries include

• D49 = T15+C9+M4+I10+O4+F2 (HLT TDR baseline)
• D60 = T15+C10+M4+I10+O4+F3 (With HFNose)
• D68 = T21+C11+M6+I11+O5+F4 (For HGCAL study on evolution of detector)
• D70 = T21+C13+M7+I11+O6+F6 (For HGCAL study on evolution of detector)
• D76 = T21+C14+M9+I13+O7+F6
• D77 = T24+C14+M9+I13+O7+F6 (Current default scenario)
• D80 = T25+C14+M9+I13+O7+F6
• D81 = T26+C14+M9+I13+O7+F6
• D82 = T21+C15+M9+I13+O7+F7
• D83 = T24+C16+M9+I13+O7+F6
• D84 = T24+C13+M7+I11+O6+F6 (For HGCAL study on evolution of HGCal replacing D70)
• D85 = T24+C14+M9+I14+O7+F6

This issue is to collect feedback, or need if other version of geometry should be preserved.

The plan is to implement this proposal when the next release series start, i.e. 12_6_X.

[cmsbuild]

A new Issue was created by @srimanob Phat Srimanobhas.

@Dr15Jones, @perrotta, @dpiparo, @makortel, @smuzaffar, @qliphy can you please review it and eventually sign/assign? Thanks.

cms-bot commands are listed here

[srimanob]

FYI @kpedro88 @bsunanda @AdrianoDee

@trtomei @fwyzard Do you still need to keep D49 for HLT?

[srimanob]

assign upgrade

[cmsbuild]

New categories assigned: upgrade

@AdrianoDee,@srimanob you have been requested to review this Pull request/Issue and eventually sign? Thanks

[srimanob]

FYI @cms-sw/simulation-l2 @cms-sw/hgcal-dpg-l2

[kpedro88]

• will need to add a new scenario with HF nose (compatible with one of the latest HGCal geometries)
• @srimanob @AdrianoDee we may want to take this opportunity to reset the workflow numbers... it's possibly been three years since we last did this in #27449

[fabiocos]

from the MTD point of view keeping just scenario I15 is ok, this will allow us to drop support in the code for obsolete scenarios (like single disc ring-shaped ETL in D49), simplifying the structure. Some initial cleaning is already proposed for BTL in #38589 (not needing this proposal), in preparation of a new BTL geometry (and MTD scenario) to be added in the coming weeks/months.

[srimanob]

• will need to add a new scenario with HF nose (compatible with one of the latest HGCal geometries)

This is handled in https://github.com/cms-sw/cmssw/pull/38596 (Thanks @bsunanda)

[srimanob]

I've made a draft PR, so ppl can check: https://github.com/cms-sw/cmssw/pull/38702

[kpedro88]

@srimanob what do you think about resetting the workflow numbers?

[srimanob]

Hi @kpedro88 Yeah, that is what I would like to ask you. If D86 will be the first workflow, should it just be 20000.0 after resetting?

Currently, the first upgrade workflow after cleaning will be 38600.0 2026D86+FourMuPt_1_200_pythia8_GenSimHLBeamSpot+DigiTrigger+RecoGlobal+HARVESTGlobal

[kpedro88]

Yes; in particular, you should change this line as follows: numWFConflict = [[25000,26000],[50000,51000]]

[srimanob]

OK. That line I updated already in the PR, but just to reflect the drop of workflows. I will update again.

[cbotta]

Hello. The L1T team would like to ask if the D49 could be kept. It can be useful for validation studies on the HLT-TDR samples (that we are using for L1T developments since a year). And I think HLT Upgrade would echo this.

[srimanob]

@cbotta Do you have an idea on how long will you need it? Could you please provide more information on the schedule of development?

The current plan is to drop by the next release series, i.e. 12_6 at the end of August.

[cbotta]

@srimanob After discussing with the team: we will very likely have our new MC production in 12_5. We think it is useful to keep the D49 up to this release, so that we have one release that can run on both MC productions. But after that we will switch to the new samples, and so it is fine to remove it from 12_6. Thanks!

[srimanob]

@cbotta This is great, thanks. I will add this bullet to tomorrow SIM meeting.

[srimanob]

Hi @cbotta @cms-sw/l1-l2 May I come back to the question on dropping geometries. It is unclear to me as we are not done yet the L1T dev, could you please comment.

FYI @bsunanda @cms-sw/simulation-l2 @cms-sw/hcal-dpg-l2

[cbotta]

Hi. My understanding is that L1T is still targeting 12_5_0 for the MC production, therefore my previous comment still holds. @cecilecaillol

[srimanob]

Hi @cbotta Thanks. The question is if we can handle the cleaning in master (12_6) now, or we should wait a bit until all L1T devs converge.

[cbotta]

If possible I would wait for @cecilecaillol to confirm that all what we need can be ported to 12_5, and we can start the production there. Should be happening soon

[srimanob]

@cbotta @cecilecaillol @cms-sw/l1-l2 After the simulation meeting today, there are requests to merge the clean up PR https://github.com/cms-sw/cmssw/pull/38702 asap. D49 will not be supported anymore in 12_6.

From your plan, it should be OK as all L1T code will be backported to 12_5, so 12_5 will be the release which support both D49 and D88. If there is an objection, please raise it here before the next release meeting (27 Sep). Thanks for understanding.

FYI @bsunanda @fabiocos

[cecilecaillol]

We should have a 12_5_1 release for our MC production so this is fine from our side

[srimanob]

Unit test failures after merging cleaning PR:

• Validation/Geometry/test/runP_HGCal_cfg.py
• Geometry/TrackerGeometryBuilder/test/python/testPixelTopologyMapTest_cfg.py
• SimTracker/TrackerMaterialAnalysis/test/trackingMaterialProducer10GeVNeutrino_ForPhaseII.py
• SLHCUpgradeSimulations/Geometry/test/phase2_digi_reco_pixelntuple_cfg.py
• SLHCUpgradeSimulations/Geometry/test/writeFile_phase2_cfg.py
• CondTools/SiPixel/test/SiPixelTemplateDBObjectUploader_Phase2_cfg.py

They call D49 or D76. This should be updated to D88 or later.

FYI @cms-sw/geometry-l2 @cms-sw/trk-dpg-l2 Should I update all to D88? If OK, I can provide the PR. Thanks.

[bsunanda]

Please use D92 (or the later version with new Tracker and MTD) and not D88.

---

From: Phat Srimanobhas @.*** Sent: 29 September 2022 09:30 To: cms-sw/cmssw Cc: Sunanda Banerjee; Mention Subject: Re: [cms-sw/cmssw] Drop of D49-D85 Phase-2 geometry (Issue #38594)

Unit test failures after merging cleaning PR:

• Validation/Geometry/test/runP_HGCal_cfg.py
• Geometry/TrackerGeometryBuilder/test/python/testPixelTopologyMapTest_cfg.py
• SimTracker/TrackerMaterialAnalysis/test/trackingMaterialProducer10GeVNeutrino_ForPhaseII.py
• SLHCUpgradeSimulations/Geometry/test/writeFile_phase2_cfg.py
• CondTools/SiPixel/test/SiPixelTemplateDBObjectUploader_Phase2_cfg.py They call D49 or D76. This should be updated to D88 or later.

FYI @cms-sw/geometry-l2https://github.com/orgs/cms-sw/teams/geometry-l2 @cms-sw/trk-dpg-l2https://github.com/orgs/cms-sw/teams/trk-dpg-l2

— Reply to this email directly, view it on GitHubhttps://github.com/cms-sw/cmssw/issues/38594#issuecomment-1261879490, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ABGMZOQQKLZA5KOVNKDQVNTWAVAQZANCNFSM52UCJQSQ. You are receiving this because you were mentioned.Message ID: @.***>

[mmusich]

@srimanob I suppose you can also clean up the T* versioned autoCond Global Tags for the Tracker geometries that have been removed?

https://github.com/cms-sw/cmssw/blob/4e89b5b212dc118b0b17abffd052144eea7ebe04/Configuration/AlCa/python/autoCondPhase2.py#L110-L119

@cms-sw/alca-l2

[srimanob]

@srimanob I suppose you can also clean up the T* versioned autoCond Global Tags for the Tracker geometries that have been removed?

https://github.com/cms-sw/cmssw/blob/4e89b5b212dc118b0b17abffd052144eea7ebe04/Configuration/AlCa/python/autoCondPhase2.py#L110-L119

@cms-sw/alca-l2

Hi @mmusich @cms-sw/alca-l2 Sure, I can include in my PR. Thanks for information.

[mmusich]

Please use D92 (or the later version with new Tracker and MTD) and not D88.

I see a couple of problems:

• in: https://github.com/cms-sw/cmssw/blob/4e89b5b212dc118b0b17abffd052144eea7ebe04/SLHCUpgradeSimulations/Geometry/test/phase2_digi_reco_pixelntuple_cfg.py#L32 as far as I can see there is no suitable GEN-SIM dataset for D92[^1]
• in: https://github.com/cms-sw/cmssw/blob/4e89b5b212dc118b0b17abffd052144eea7ebe04/CondTools/SiPixel/test/createTestDBObjects.sh#L20 we don't have the suitable mapping files in https://github.com/cms-sw/cmssw/tree/master/CondTools/SiPixel/data nor in https://github.com/cms-data/CalibTracker-SiPixelESProducers (@tvami FYI)

[^1]: from DAS $ dasgoclient -query='dataset dataset=/RelValSingleMuPt10/*2026D92*/GEN-SIM' | wc -l is 0

[srimanob]

Let's go for D88 in unit tests until we validate D92 and move it to baseline, and have everything ready.

[mmusich]

@srimanob

PR https://github.com/cms-sw/cmssw/pull/39535 should take care of these:

Geometry/TrackerGeometryBuilder/test/python/testPixelTopologyMapTest_cfg.py SimTracker/TrackerMaterialAnalysis/test/trackingMaterialProducer10GeVNeutrino_ForPhaseII.py SLHCUpgradeSimulations/Geometry/test/phase2_digi_reco_pixelntuple_cfg.py SLHCUpgradeSimulations/Geometry/test/writeFile_phase2_cfg.py

as for the others:

Validation/Geometry/test/runP_HGCal_cfg.py

I would let it to HGCAL experts (FYI @cms-sw/hgcal-dpg-l2 )

while

CondTools/SiPixel/test/SiPixelTemplateDBObjectUploader_Phase2_cfg.py

requires more involved changes, including an update of the cms-data externals, that I would leave to the Pixel CPE experts (@tvami @SanjanaSekhar)