New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[DD4he] Disable DD4hep workflow 11634.911 in PR tests while it produces unstable results #33568
Conversation
+code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-33568/22362
|
A new Pull Request was created by @cvuosalo (Carl Vuosalo) for master. It involves the following packages: Configuration/PyReleaseValidation @jordan-martins, @chayanit, @wajidalikhan, @kpedro88, @cmsbuild, @srimanob can you please review it and eventually sign? Thanks. cms-bot commands are listed here |
please test |
please test |
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-0dd385/14686/summary.html Comparison SummarySummary:
|
+1 |
merge |
The DD4hep workflow 11634.911 is once again showing instability in its results. For several weeks it was stable, but some recent change, possibly the new versions of DD4hep and ROOT, has re-activated its instability. Any change in the geometry causes a change in the simulation histories, which then changes the bit-wise results of simulation, even if there are no qualitative changes. Somehow, from run to run of the workflow, the geometry can be slightly changed, possibly due to an uninitialized variable taking on different values each time.
Note that the instability being observed is on the 10% level, that is, about 90% of the time the workflow produces consistent results.
The failed comparisons in PR tests for this workflow create unnecessary confusion, so it is best to remove the workflow for now until the instability can be resolved.