cms-sw
fix lock contention in source (14_0_X backport)
PR description:
Fixes lock contention in source. Lock is used for setting up waiting on the condition variable, but it should be released after. However, the lock is automatically reacquired at the end of wait as described: https://en.cppreference.com/w/cpp/thread/condition_variable/wait
In production HLT this is seen to take significant time in the source (8-20% depending on data flow conditions and configuration) and could be the cause preventing using top ~10% of the CPU capacity in HLT.
Also, more logging (destination info) is added for the case of file broker connection issues.
PR validation:
Tested in daq3val system with both FedRawDataInputSource and DAQSource.
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:
Backport of https://github.com/cms-sw/cmssw/pull/45097 for 14_0_X release cycle. Reason: performance bugfix in DAQ/HLT
A new Pull Request was created by @smorovic for CMSSW_14_0_X.
It involves the following packages:
- EventFilter/Utilities (daq)
@smorovic, @emeschi, @cmsbuild can you please review it and eventually sign? Thanks. @missirol, @Martin-Grunewald this is something you requested to watch as well. @rappoccio, @sextonkennedy, @antoniovilela you are the release manager for this.
cms-bot commands are listed here
+1
Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-f30099/39626/summary.html
COMMIT: 5c7f19b3c1a6d104f50bdbf30c215f6d52eb6c14
CMSSW: CMSSW_14_0_X_2024-05-30-1100/el8_amd64_gcc12
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week1/cms-sw/cmssw/45098/39626/install.sh to create a dev area with all the needed externals and cmssw changes.
Comparison Summary
Summary:
- You potentially added 63 lines to the logs
- Reco comparison results: 58 differences found in the comparisons
- DQMHistoTests: Total files compared: 48
- DQMHistoTests: Total histograms compared: 3338712
- DQMHistoTests: Total failures: 1205
- DQMHistoTests: Total nulls: 0
- DQMHistoTests: Total successes: 3337487
- DQMHistoTests: Total skipped: 20
- DQMHistoTests: Total Missing objects: 0
- DQMHistoSizes: Histogram memory added: 0.0 KiB( 47 files compared)
- Checked 202 log files, 165 edm output root files, 48 DQM output files
- TriggerResults: no differences found
This pull request is fully signed and it will be integrated in one of the next CMSSW_14_0_X IBs (tests are also fine) and once validation in the development release cycle CMSSW_14_1_X is complete. This pull request will now be reviewed by the release team before it's merged. @antoniovilela, @rappoccio, @sextonkennedy (and backports should be raised in the release meeting by the corresponding L2)
@cms-sw/orp-l2
Potentially this could improve HLT performance inefficiency (as seen in throughput tests where we try to saturate HLT farm and notice source spending time on this lock). After this is merged in 14_0_X, we would like to ask for a patch release (or a full release if that is scheduled in the next days) that will include this PR as well as: https://github.com/cms-sw/cmssw/pull/44980
