Replace trackingPhase1PU70 workflows with CA seeding #16911
Conversation
|
@cmsbuild, please test |
|
The tests are being triggered in jenkins. |
|
A new Pull Request was created by @makortel (Matti Kortelainen) for CMSSW_9_0_X. It involves the following packages: Configuration/Eras @dmitrijus, @cvuosalo, @franzoni, @fabozzi, @cmsbuild, @srimanob, @slava77, @vanbesien, @hengne, @davidlange6 can you please review it and eventually sign? Thanks. cms-bot commands are listed here #13028 |
|
@felicepantaleo Could you double-check the CA parameters? |
|
@makortel looks good to me |
|
Comparison job queued. |
|
Comparison is ready The workflows 1003.0, 1001.0, 1000.0, 140.53, 136.731, 4.22 have different files in step1_dasquery.log than the ones found in the baseline. You may want to check and retrigger the tests if necessary. You can check it in the "files" directory in the results of the comparisons |
|
if #16885 is merged on Friday, this will likely need a rebase |
|
@makortel |
|
Sure, is there preference to do it now or wait an IB? |
|
do it now, I'll wait for IB to test |
Configurations will be cleaned up later.
… special workflows
cmsbuild
added
comparison-pending
tests-pending
and removed
comparison-available
tests-approved
labels
|
Pull request #16911 was updated. @dmitrijus, @cvuosalo, @franzoni, @fabozzi, @cmsbuild, @srimanob, @slava77, @vanbesien, @hengne, @davidlange6 can you please check and sign again. |
|
@cmsbuild, please test Following the discussion above, I switched the "old" phase1 seeding back to the default. So essentially this PR now replaced |
|
The tests are being triggered in jenkins. |
makortel
changed the title
|
Comparison job queued. |
|
Comparison is ready @slava77 comparisons for the following workflows were not done due to missing matrix map:
|
This PR switches the default phase1 pixel seeding to cellular automaton (tuning by @felicepantaleo). The workflows ofThis PR replaces thetrackingPhase1PU70are replaced with workflows of the previous seeding (trackingPhase1QuadProp, suggestions for better name are welcome) to be kept as a reference for some time.trackingPhase1PU70workflows with CA seeding (tuning by @felicepantaleo). The goal is to switch the default phase1 tracking to CA seeding once the duplicate rate increase is better understood (and fixed/mitigated), therefore the "CA seeding" is already implemented undertrackingPhase1sub-era, and the default trackign undertrackingPhase1QuadPropera (suggestions for better name are welcome). Full cleanup oftrackingPhase1PU70will come later (after the pre2 dust settles).Here are MTV plots for 1000 ttbar+35PU events in 900pre1 (using 810pre16 DIGI-RAW as an input) comparing the CA seeding to the default.
https://mkortela.web.cern.ch/mkortela/tracking/validation/CMSSW_9_0_0_pre1_caseeding/
Timing decreases by ~5 % for the full trackingOnly reco job (becomes ~10 % when excluding most expensive other-than-iterative-tracking modules, mostly validation).
The previous seeding is kept workflows 10024.[45] (full/trackingOnly), so there should be no changes between old 10024.[01] and new 10024.[45].To minimize merge conflicts the developments are based on @ebrondol's #16858 (in the hope of getting this in pre2).Tested in CMSSW_9_0_X_2016-11-30-2300 (rebased on top of CMSSW_9_0_X after IB CMSSW_9_0_X_2016-12-09-1100 and before the next). No changes expected in standard workflows (special 10024.[45] show the changes from
trackingPhase1PU70to CA).for phase0, phase1 changes are described above.and for phase2 there are changes because of #16858 (but no changes on top of that).@rovere @VinInn @felicepantaleo @ebrondol @mtosi