Fake DQMHistosTests failures in wf 11634.0 and 12434.0 (Hcal/DigiTask/LETDC*)

[silviodonato]

@abdoulline after having integrated #29035 (CMSSW_11_1_X_2020-02-27-2300), we started to have fake DQMHistosTests failures in wf 11634.0 and 12434.0 (Hcal/DigiTask/LETDCTime, LETDCTimevsADC, LETDCvsADC, and LETDCvsTS)

For instance:
https://tinyurl.com/txjqbx9

Even if the two distributions match perfectly, there is a large difference in the number of entries (260 vs 240) and on the mean. I guess the difference is due to the overflow bin.

Other examples:
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-e25e49/4951/summary.html
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-4bcc52/4945/summary.html
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-88bcf8/4928/summary.html
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-233b40/4930/summary.html
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-a54065/4933/summary.html
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-0b1657/4936/summary.html
Click on "DQM bin by bin comparison" to see the difference.

If I remember correctly, a possible explanation is that during the IB test we run multithreading while in the PR test we run single thread. Anyway, this should not happen.

@abdoulline could you have a look to this issue?

cc: @mrodozov

[silviodonato]

assign simulation

[cmsbuild]

New categories assigned: simulation

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

[cmsbuild]

A new Issue was created by @silviodonato Silvio Donato.

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

cms-bot commands are listed here

[abdoulline]

Take the liberty to include HCAL DQM colleagues
@DryRun . @lwang046
There is something problematic in DQM/HcalTasks "DigiTask" with the histos (variables initialization?) related to TDC, as there was already this sort of issues in the past -
#28367

[abdoulline]

Or (as David Yu @DryRun supposed elsewhere)
there could be some TDC simulation (initialization) issue, which we've hidden previously by excluding TDC simulation for 2017 era (as a followup of issue #28367) and which now became "visible" in Run3 wf's with #29035 integrated.
Asked Long Wang @lwang046 to take a close look at the things with wf 11634.0

[silviodonato]

solved by #29076

[Dr15Jones]

@silviodonato The pull request you referenced as solving this issue is just this issue.

[abdoulline]

This is the link to the fix #29083

[kpedro88]

is this completely fixed now? if so, the issue should be closed

[DryRun]

Maybe @lwang046 or @abdoulline should confirm, but I believe #28368 fixed the issue once and for all.

[abdoulline]

I believe so.

…

On Wed, 15 Apr 2020, David Yu wrote: Maybe @lwang046 or @abdoulline should confirm, but I believe #28368 fixed the issue once and for all. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, orunsubscribe.[ABGHJWR4HSPXRERQ2FLQR5LRMXEYNA5CNFSM4K72GMXKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXH JKTDN5WW2ZLOORPWSZGOESNF7UY.gif]

[abdoulline]

@DryRun
in fact, not it's not #28368, but #29083 which fixed the issue once for all.