RelVals 546.0, 547.0, 548.0 failing with ProductNotFound

[iarspider]

RelVals 546.0, 547.0, 548.0 is failing for all architectures with ProductNotFound error:

----- Begin Fatal Exception 09-Apr-2024 05:27:42 CEST-----------------------
An exception of category 'ProductNotFound' occurred while
   [0] Processing  Event run: 1 lumi: 1 event: 2 stream: 1
   [1] Running path 'nanoAOD_step'
   [2] Calling method for module GenPartIsoProducer/'genIso'
Exception Message:
Principal::getByToken: Found zero products matching all criteria
Looking for type: std::vector<reco::GenParticle>
Looking for module label: finalGenParticles
Looking for productInstanceName: 

   Additional Info:
      [a] If you wish to continue processing events after a ProductNotFound exception,
add "TryToContinue = cms.untracked.vstring('ProductNotFound')" to the "options" PSet in the configuration.

----- End Fatal Exception -------------------------------------------------

Probably caused by #44623

[iarspider]

assign xpog

[cmsbuild]

New categories assigned: xpog

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

[cmsbuild]

cms-bot internal usage

[cmsbuild]

A new Issue was created by @iarspider.

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

cms-bot commands are listed here

[vlimant]

related to #44623; I'll take a look but please @bonanomi do too

[bonanomi]

Hi @vlimant , I did a comparison between the code I pushed and the one I ran locally (successfully) and I did not spot any difference.
I am also puzzled by Looking for module label: finalGenParticles, since this is already defined in PhysicsTools/NanoAOD/python/genparticles_cff.py

[vlimant]

some like

diff --git a/PhysicsTools/NanoAOD/python/nanogen_cff.py b/PhysicsTools/NanoAOD/python/nanogen_cff.py
index 437d3869a37..45a8059ee40 100644
--- a/PhysicsTools/NanoAOD/python/nanogen_cff.py
+++ b/PhysicsTools/NanoAOD/python/nanogen_cff.py
@@ -98,6 +98,8 @@ def customizeNanoGEN(process):
     process.nanogenSequence.remove(process.genParticles2HepMCHiggsVtx)
     process.nanogenSequence.remove(process.genParticles2HepMC)
     process.nanogenSequence.remove(process.mergedGenParticles)
+    process.nanogenSequence.remove(process.genIso)
+    delattr(process.genParticleTable.externalVariables,"iso")
     nanoGenCommonCustomize(process)
     return process

seems to make it work, as a hack ; i.e removing genIso when NANOGEN is ran (BTW https://gitlab.cern.ch/cms-nanoAOD/xpog-coordination/-/issues/78)

[vlimant]

maybe there is something more intelligent to do though : if you can think of something @bonanomi please do.
otherwise please make a 14.1 PR and include the same "fix" in #44654 (so that we do not brake 14.0)

[hqucms]

@bonanomi @namapane Could you please implement the fix for both master and the 14_0_X backport #44654?

[namapane]

@hqucms I'm going to check for a more intelligent solution as metioned by @vlimant, but it will take a moment to test. If this is critical we could go for the simple solution proposed above (skip the module in NANOGEN). Please let me know.

[hqucms]

I think it's better to put the simple fix first, then you can always update it with a better solution later :)