Proper use of eras for L1Reco and a fastSim fix

[Dr15Jones]

The tests in CMSSW_8_1_DEVEL_X are failing because of an unrunnable schedule when using fastSim with the legacy L1 trigger. Although it would have been possible to fix just that problem, the L1TReco_cff file was not using the modern recommended practice for eras. Therefore I also updated the use of eras.

[cmsbuild]

A new Pull Request was created by @Dr15Jones (Chris Jones) for CMSSW_8_1_X.

It involves the following packages:

L1Trigger/Configuration
L1Trigger/L1ExtraFromDigis

@cmsbuild, @rekovic, @mulhearn, @davidlange6 can you please review it and eventually sign? Thanks.
@Martin-Grunewald this is something you requested to watch as well.
@slava77, @smuzaffar you are the release manager for this.

cms-bot commands are list here #13028

[Dr15Jones]

I used the following program to test that my changes give functionally equivalent configurations

import sys
import FWCore.ParameterSet.Config as cms

from Configuration.Eras.Modifier_stage1L1Trigger_cff import stage1L1Trigger
from Configuration.Eras.Modifier_stage2L1Trigger_cff import stage2L1Trigger
from Configuration.Eras.Modifier_fastSim_cff import fastSim

modifiers = []
for i in sys.argv[1:]:
    print "***********"
    print i
    print "***********"
    if i == "fastSim":
        modifiers.append(fastSim)
    if i == "stage1":
        modifiers.append(stage1L1Trigger)
    if i == "stage2":
        modifiers.append(stage2L1Trigger)


process = cms.Process("Test",*modifiers)
process.load("L1Trigger.Configuration.L1TReco_cff")

print process.dumpPython()

I applied this to both the original version of the code and this pull request for the following combinations of era

1. no eras
2. just fastSim
3. stage1
4. stage1 and fastSim
5. stage2
6. stage2 and fastSim

Doing a diff on the output showed that the change I made gave functionally equivalent configurations. By 'functinally equivalent' I mean

1. the new configuration may contain unuse module declarations and additional empty sequences
2. the change when running fastSim with legacy where l1L1GtObjectMap is removed

[Dr15Jones]

please test

[cmsbuild]

The tests are being triggered in jenkins.
https://cmssdt.cern.ch/jenkins/job/ib-any-integration/15573/console

[cmsbuild]

-1

Tested at: c40cda6

You can see the results of the tests here:
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-16118/15573/summary.html

I found follow errors while testing this PR

Failed tests: UnitTests

• Unit Tests:

I found errors in the following unit tests:

[cmsbuild]

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-16118/15573/summary.html

[Dr15Jones]

please test

[cmsbuild]

The tests are being triggered in jenkins.
https://cmssdt.cern.ch/jenkins/job/ib-any-integration/15584/console

[cmsbuild]

+1
Tested at: c40cda6
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-16118/15584/summary.html

[cmsbuild]

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-16118/15584/summary.html

[Dr15Jones]

Ping
This is needed for the next phase of the threaded framework.

[Dr15Jones]

I looked through all the comparison plots and the only one with even a minor difference was

25202.0 HLT >> Muon >> Distributions >> HLT_IsoTkMu18
https://cmssdt.cern.ch/SDT/jenkins-artifacts/baseLineComparisons/CMSSW_8_1_X_2016-10-06-1100+16118/16251/25202.0_TTbar_13+TTbar_13+DIGIUP15_PU25+RECOUP15_PU25+HARVESTUP15_PU25/HLT_Muon_Distributions_HLT_IsoTkMu18.html

@slava77 from your experience with the comparison plots, is this an actual item to be worried about?

[slava77]

On 10/7/16 7:09 AM, Chris Jones wrote:

I looked through all the comparison plots and the only one with even a
minor difference was

25202.0 HLT >> Muon >> Distributions >> HLT_IsoTkMu18
https://cmssdt.cern.ch/SDT/jenkins-artifacts/baseLineComparisons/CMSSW_8_1_X_2016-10-06-1100+16118/16251/25202.0_TTbar_13+TTbar_13+DIGIUP15_PU25+RECOUP15_PU25+HARVESTUP15_PU25/HLT_Muon_Distributions_HLT_IsoTkMu18.html

HLT_IsoTkMu was on our ignore list for probably 6 months.
Comparisons for it regularly show random differences.

@slava77 https://github.com/slava77 from your experience with the
comparison plots, is this an actual item to be worried about?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#16118 (comment), or
mute the thread
https://github.com/notifications/unsubscribe-auth/AEdcbjXbXeCQnESZSgT_lcc_ryQO1vryks5qxlKUgaJpZM4KQIv6.