Repository navigation
Conversation
Add TRD to the cluster types requested by the global track/cluster reader feeding the TPC TimeSeries task. The TimeSeries workflow consumes ITS-TPC-TRD-TOF tracks and therefore requests TRD calibrated tracklets. Without TRD in --cluster-types, the workflow has no producer for TRD/CTRACKLETS/0 and fails workflow validation before execution. Keep the existing TimeSeries track-source selection and all other MC workflow settings unchanged. Validation: - Python syntax check passed - functional change limited to adding TRD to the TimeSeries reader --cluster-types list
|
REQUEST FOR PRODUCTION RELEASES: This will add The following labels are available |
|
Reference JIRA: |
|
Hello @sawenzel and @jackal1-66, Could you please approve and merge the modification? It is a single line that enables TRD tracklets in the MC production for the time series. We need it for: It is blocking time series productions. |
|
Hi @jackal1-66 and @sawenzel, |
|
Hi @jackal1-66 and @sawenzel, Could you please have a look, and if you agree, merge? It is a one-line change: adding TRD to the input. Thanks! |
|
I am testing an alternative approach since we should not add detector names statically. This will fail immediately, the moment TRD is not part of a data-taking run. Rather I propose to follow up on #2141 and make the input detector list fully dynamic. Testing it right now. |
sawenzel
left a comment
There was a problem hiding this comment.
I propose to hold this PR and follow up with a more flexible/dynamic path.
This makes the TPC time-series task in the MC workflow take its inputs from the active detectors. - The reader used fixed cluster types without TRD, so the time series failed with "No matching output found for TRD/CTRACKLETS/0" once it requested TRD tracklets. - Track and cluster types now come from cleanDetectorInputList. - The same track list goes to o2-tpc-time-series-workflow via --track-sources, if the async software supports it. - TOF and TRD are optional, and the task requires ITS, TPC and FT0. https://its.cern.ch/jira/browse/O2-7264 AliceO2Group#2532 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
Do you expect my one-line fix to fail for the specific production currently blocked in O2-7264? We have a production under preparation blocked now: If the dynamic solution is not yet ready for production, I suggest enabling TRD with this minimal fix for the current production and replacing it with the general dynamic solution afterwards. Is your dynamic solution ready to be used for this production now? Please give me a short answer: yes or no. @alcaliva , @catalinristea - when do you plan to start MC production O2-7264 ? I assume preparation is almost ready? |
This fixes the TRD/CTRACKLETS failure of the TPC time-series task in the MC workflow and makes its inputs follow the active detectors. - The reader used fixed cluster types without TRD, so the time series failed with "No matching output found for TRD/CTRACKLETS/0" once it requested TRD tracklets. - Track types now follow the active detectors and the anchored TRD and TOF matching sources. - The same track list goes to o2-tpc-time-series-workflow via --track-sources. - Builds with their own readers get --disable-root-input, so only one reader list exists. - TOF and TRD are optional, and the task requires ITS, TPC and FT0. - Builds without --track-sources keep the previous command and require TOF and TRD. https://its.cern.ch/jira/browse/O2-7264 AliceO2Group#2532 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This fixes the TRD/CTRACKLETS failure of the TPC time-series task in the MC workflow and makes its inputs follow the active detectors. - The reader used fixed cluster types without TRD, so the time series failed with "No matching output found for TRD/CTRACKLETS/0" once it requested TRD tracklets. - Track types now follow the active detectors and the anchored TRD and TOF matching sources. - The same track list goes to o2-tpc-time-series-workflow via --track-sources. - Builds with their own readers get --disable-root-input, so only one reader list exists. - TOF and TRD are optional, and the task requires ITS, TPC and FT0. - Builds without --track-sources keep the previous command and require TOF and TRD. https://its.cern.ch/jira/browse/O2-7264 #2532 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
I merged #2141. A new tag will be prepared if tests on tomorrows daily succeed. |
The anchored-MC TPC TimeSeries task consumes TRD-matched tracks and
requests the TRD calibrated tracklet input TRD/CTRACKLETS/0.
However, the o2-global-track-cluster-reader feeding the TimeSeries task
was configured with:
--cluster-types "FT0,TOF,TPC"
so no producer for TRD/CTRACKLETS/0 was present in the generated workflow.
This caused workflow validation to fail with:
[ERROR] invalid workflow in o2-tpc-time-series-workflow:
No matching output found for TRD/CTRACKLETS/0
as requested by data processor "tpc-time-series".
The fix is deliberately minimal:
--cluster-types "FT0,TOF,TPC"
->
--cluster-types "FT0,TOF,TPC,TRD"
No track-source selection or other MC workflow configuration is changed.
This affects anchored MC productions with TPC TimeSeries enabled through
ADDTIMESERIESINMC.
Validation:
configuration.
Branch: miranov25:mcfix
Commit: 9936e59