Closed mbluj closed 1 week ago
cms-bot internal usage
+code-checks
Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-46679/42606
A new Pull Request was created by @mbluj for master.
It involves the following packages:
@cmsbuild, @ftorrresd, @hqucms can you please review it and eventually sign? Thanks. @AnnikaStein, @gpetruc this is something you requested to watch as well. @antoniovilela, @mandrenguyen, @rappoccio, @sextonkennedy you are the release manager for this.
cms-bot commands are listed here
enable nano
please test
+1
Size: This PR adds an extra 12KB to repository
Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-8948f5/42772/summary.html
COMMIT: 76197dbbfb0948eae3bab52b4fa2916fd6db432a
CMSSW: CMSSW_14_2_X_2024-11-12-1100/el8_amd64_gcc12
Additional Tests: NANO
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week1/cms-sw/cmssw/46679/42772/install.sh
to create a dev area with all the needed externals and cmssw changes.
There are some workflows for which there are errors in the baseline: 2024.101001 step 1 2024.202001 step 1 2024.303001 step 1 The results for the comparisons for these workflows could be incomplete This means most likely that the IB is having errors in the relvals.The error does NOT come from this pull request
Summary:
Summary:
Nano size comparison Summary:
Sample | kb/ev | ref kb/ev | diff kb/ev | ev/s/thd | ref ev/s/thd | diff rate | mem/thd | ref mem/thd |
---|---|---|---|---|---|---|---|---|
2500.001 | 3.053 | 3.052 | 0.001 ( +0.0% ) | 2.71 | 2.71 | +0.1% | 2.520 | 2.523 |
2500.002 | 3.165 | 3.164 | 0.001 ( +0.0% ) | 2.43 | 2.45 | -0.8% | 2.943 | 2.953 |
2500.003 | 3.107 | 3.105 | 0.002 ( +0.1% ) | 2.54 | 2.56 | -0.9% | 2.931 | 2.933 |
2500.011 | 1.621 | 1.621 | 0.000 ( +0.0% ) | 4.14 | 4.22 | -1.9% | 2.609 | 2.611 |
2500.012 | 2.153 | 2.153 | 0.000 ( +0.0% ) | 2.53 | 2.55 | -0.6% | 2.800 | 2.800 |
2500.013 | 1.975 | 1.975 | 0.000 ( +0.0% ) | 3.46 | 3.55 | -2.5% | 2.708 | 2.718 |
2500.021 | 0.022 | 0.022 | 0.000 ( +0.0% ) | 0.86 | 0.86 | -0.4% | 2.560 | 2.555 |
2500.022 | 0.022 | 0.022 | 0.000 ( +0.0% ) | 0.81 | 0.83 | -3.0% | 2.557 | 2.564 |
2500.023 | 0.022 | 0.022 | 0.000 ( +0.0% ) | 0.81 | 0.83 | -1.5% | 2.411 | 2.421 |
2500.024 | 0.022 | 0.022 | 0.000 ( +0.0% ) | 0.63 | 0.63 | -1.0% | 2.652 | 2.637 |
2500.031 | 0.035 | 0.035 | 0.000 ( +0.0% ) | 0.75 | 0.75 | +0.5% | 2.627 | 2.633 |
2500.032 | 0.036 | 0.036 | 0.000 ( +0.0% ) | 0.75 | 0.76 | -1.6% | 2.590 | 2.583 |
2500.033 | 0.037 | 0.037 | 0.000 ( +0.0% ) | 0.69 | 0.71 | -2.9% | 2.661 | 2.673 |
2500.034 | 0.036 | 0.036 | 0.000 ( +0.0% ) | 0.70 | 0.70 | -0.4% | 2.656 | 2.651 |
2500.101 | 2.791 | 2.788 | 0.004 ( +0.1% ) | 6.52 | 6.71 | -2.9% | 2.614 | 2.616 |
2500.111 | 1.447 | 1.447 | 0.000 ( +0.0% ) | 13.03 | 13.27 | -1.8% | 2.323 | 2.320 |
2500.112 | 1.867 | 1.867 | 0.000 ( +0.0% ) | 11.09 | 11.08 | +0.1% | 2.397 | 2.395 |
2500.131 | 0.747 | 0.747 | 0.000 ( +0.0% ) | 17.08 | 17.44 | -2.1% | 1.514 | 1.520 |
2500.201 | 2.618 | 2.617 | 0.001 ( +0.0% ) | 5.54 | 5.57 | -0.5% | 2.171 | 2.171 |
2500.211 | 1.786 | 1.786 | 0.000 ( +0.0% ) | 11.30 | 11.43 | -1.1% | 2.384 | 2.389 |
2500.212 | 2.182 | 2.182 | 0.000 ( +0.0% ) | 9.58 | 9.97 | -3.9% | 2.473 | 2.473 |
2500.221 | 2.030 | 2.028 | 0.002 ( +0.1% ) | 5.86 | 5.93 | -1.1% | 2.099 | 2.095 |
2500.222 | 3.417 | 3.415 | 0.002 ( +0.1% ) | 5.48 | 5.53 | -1.0% | 2.194 | 2.188 |
2500.223 | 9.382 | 9.380 | 0.002 ( +0.0% ) | 1.69 | 1.70 | -0.7% | 2.266 | 2.263 |
2500.224 | 6.066 | 6.063 | 0.002 ( +0.0% ) | 0.57 | 0.57 | -0.8% | 2.206 | 2.206 |
2500.225 | 6.113 | 6.111 | 0.002 ( +0.0% ) | 0.53 | 0.54 | -0.6% | 2.424 | 2.424 |
2500.226 | 3.111 | 3.109 | 0.002 ( +0.1% ) | 5.61 | 5.75 | -2.6% | 2.188 | 2.183 |
2500.227 | 1.437 | 1.437 | 0.000 ( +0.0% ) | 11.09 | 11.22 | -1.1% | 1.431 | 1.433 |
2500.231 | 1.449 | 1.449 | 0.000 ( +0.0% ) | 9.36 | 9.58 | -2.3% | 2.278 | 2.285 |
2500.232 | 2.441 | 2.441 | 0.000 ( +0.0% ) | 8.88 | 9.17 | -3.2% | 2.382 | 2.384 |
2500.233 | 4.933 | 4.933 | 0.000 ( +0.0% ) | 2.51 | 2.52 | -0.5% | 2.452 | 2.445 |
2500.234 | 3.700 | 3.700 | 0.000 ( +0.0% ) | 0.72 | 0.73 | -1.0% | 2.174 | 2.174 |
2500.235 | 3.730 | 3.730 | 0.000 ( +0.0% ) | 0.69 | 0.70 | -1.3% | 2.374 | 2.368 |
2500.236 | 2.232 | 2.232 | 0.000 ( +0.0% ) | 9.48 | 9.51 | -0.3% | 2.385 | 2.381 |
2500.237 | 1.016 | 1.016 | 0.000 ( +0.0% ) | 16.26 | 16.31 | -0.3% | 1.461 | 1.440 |
2500.241 | 9.404 | 9.404 | 0.000 ( +0.0% ) | 3.43 | 3.42 | +0.0% | 1.926 | 1.914 |
2500.242 | 10.331 | 10.331 | 0.000 ( +0.0% ) | 0.83 | 0.82 | +0.9% | 1.711 | 1.274 |
2500.243 | 2.712 | 2.712 | 0.000 ( +0.0% ) | 7.67 | 7.19 | +6.6% | 1.083 | 1.077 |
2500.244 | 485.976 | 485.976 | 0.000 ( +0.0% ) | 0.53 | 0.53 | -1.0% | 1.681 | 1.677 |
2500.245 | 823.202 | 823.202 | 0.000 ( +0.0% ) | 0.69 | 0.71 | -2.1% | 1.663 | 1.686 |
2500.901 | 1.777 | 1.777 | 0.000 ( +0.0% ) | 19.58 | 19.57 | +0.0% | 1.419 | 1.418 |
2500.902 | 1.626 | 1.626 | 0.000 ( +0.0% ) | 19.40 | 20.41 | -4.9% | 1.325 | 1.323 |
2500.911 | 13.995 | 13.995 | 0.000 ( +0.0% ) | 1.98 | 3.30 | -40.0% | 1.091 | 1.083 |
2500.912 | 0.199 | 0.199 | 0.000 ( +0.0% ) | 1.18 | 1.31 | -9.9% | 0.972 | 0.978 |
2500.913 | 0.110 | 0.110 | 0.000 ( +0.0% ) | 1.13 | 1.10 | +3.3% | 0.980 | 0.977 |
+1
This pull request is fully signed and it will be integrated in one of the next master IBs (tests are also fine). This pull request will now be reviewed by the release team before it's merged. @antoniovilela, @sextonkennedy, @rappoccio, @mandrenguyen (and backports should be raised in the release meeting by the corresponding L2)
+1
PR description:
This PR fixes
TauSpinnerTableProducer
(introduced in #45828) against an issue caused by samples with events containing the H->tau,tau (HTT) decay mixed with events without this decay. Examples of such samples were discussed in https://cms-talk.web.cern.ch/t/higgs-in-wz-amcatnlofxfx-samples/55138.The nanoAOD production crashes in such samples as the
TauSpinnerTableProducer
produces a non-empty table for events with the HTT decay while empty one for events without it which leads to an inconsistent output across different events which then confuses the output module.The fix consists of producing tables with default weights for events without HTT ensuring consistency, but for price of producing not needed numbers. To be noted that similar strategy is employed in case of HTXS tables. Overhead due to this should be limited as tables with constant values are efficiently compressed.
PR validation:
Validated by processing samples with events containing the HTT decay, without the HTT decay and mixed with and without.
If this PR is a backport please specify the original PR and why you need to backport that PR. If this PR will be backported please specify to which release cycle the backport is meant for:
No backport expected.