Choose timezone
Your profile timezone:
Offline Calibration Readiness
Description
Title: Offline Calibration Readiness
Description: Offline Calibration Readiness - F.Carminati
Community: ALICE
Meeting Access Information:
- Meeting URL
http://evo.caltech.edu/evoNext/koala.jnlp?meeting=eiese9vuv9aeanIiatIi
- Phone Bridge
ID: 12 9914
---------------1-
Central European Summer Time (+0200)
Start 2012-03-29 08:30
End 2012-03-29 11:00
UNTIL
---------------39-
Central European Time (+0100)
Start 2012-12-20 08:30
End 2012-12-20 11:00
***********************************************************
Minutes of the meeting on 06 December 2012
***********************************************************
Present: Marian, Chiara, Ruben, Raffaele, Costin, Gustavo; Davide, Raphaelle, Massimo joint in the second part of the meeting
1) LHC12i processing going on smoothly
1a) run 193301 has only 10% of the statistics processed, since there were no kCalibBarrel triggers and the downscaling to 10% was applied
1a1) this hold be only a back-up solution, to be use manually if needed
1a2) Costin will remove it from production
1b) run 193047 was not processed since for some reason the MonALISA Shutle page did not update the status of the TRD pp, which was in DONE
1b1) Costin will implement in his selection a cross-check between the info in the logbook and MonALISA for the preprocessors' statuses
1b2) the run will be submitted manually by Costin
1c) Recent runs with small statistics
1c1) detector experts should be careful when evaluating a period, since in case of negative returning codes (typical example is low statistics) they are not notified and CPass1 starts (since we use the validation codes now)
2) LHC12h
2a) Marian noticed 3 runs missing from his scripts:
2a1) 189699 --> failed in the snapshot (already reported at the beginning of November)
2a1a) Marian will add the check of the snapshot to his scripts
2a2) 192497, 511 --> to be understood
2a3) The runs won't be submitted (if it is found that it was needed), since LHC12h has already enough statistics
3) Merge-by-component
3a) Costin will resubmit the test with the modified executable
3b) Chiara will commit the changes in the executables to the trunk, as soon as she can commit (problems with svn)
3c) exit code: for the time being Costin will add that as soon as any aliroot process fails (even one partial merging), the whole process will be interrupted; the result should be equivalent to the current behavior
4) New TPC modules
4a) To be enabled, a change is needed in the AddTaskTPCcalib macro
4a1) Since Peter would like not to have any new revisions, we'll use it as input file, taking it from the trunk
4a1a) Marian will check with Mikolaj that it is ok
5) HMPID still out of CPass reconstruction
5a) this affects both CPass0 and CPass1
5b) Marian will add a comment to savannah asking Giacomo what the status is, since it would be important to enable the redo in CPass1 to have HMPID in the QA
6) TPC/Calib/Correction file used as input file in some CPass processed
6a) it will be simply taken as the other OCDB files from the snapshot; if problems come, we'll go back to the current solution (no specific tests could be done so far)
7) pA triggers
7a) in pA, no SPI will be used, so we have to find out the minimum number of events to have calibration for a run which lasted ~15 minutes (which is the criteria used now, to determine the SPI rate in pp), taking into account that:
7a1) The minimum number of SPI events that the TRD seems to need to have enough statistics is ~20000
7a2) The multiplicity in pA should be a factor 4 higher that in pp; we should anyway use factor 3, to be on the safe side
7b) in "min-bias" like mode (when all will be MB trigger) we'll downscale at redo level as we did for the three runs in LHC12c that did not have kCalibBarrel, and as was done last year; again, it is needed the minimum number of events
8) Wiping of CPass output
8a) proposal to remove all the chunk-level files, and keep only the merged one; the wiping procedure should occur after the PPass of the period under consideration
8b) Detectors are asked to prepare a statement for next week
9) LHC12h: calibration readiness for VPass
9a) Mean Vertex ready
9b) while checking the Mean Vertex it was found that the MeanVertexSPD object produced online is not ok for some runs
9b1) under investigation, also in terms of consequences for VPass
9c) TRD: waiting for QA feedback on 192707-708
9d) should be ready by the end of the week
9e) MUON alignment update will be postponed after VPass if everyone else is ready, to save time
9f) Some runs were with very little statistics (between 192585 and 192676): suggestion to mark them as bad
9g) Missing detectors contacted at the end of the meeting by Chiara
Minutes of the meeting on 06 December 2012
***********************************************************
Present: Marian, Chiara, Ruben, Raffaele, Costin, Gustavo; Davide, Raphaelle, Massimo joint in the second part of the meeting
1) LHC12i processing going on smoothly
1a) run 193301 has only 10% of the statistics processed, since there were no kCalibBarrel triggers and the downscaling to 10% was applied
1a1) this hold be only a back-up solution, to be use manually if needed
1a2) Costin will remove it from production
1b) run 193047 was not processed since for some reason the MonALISA Shutle page did not update the status of the TRD pp, which was in DONE
1b1) Costin will implement in his selection a cross-check between the info in the logbook and MonALISA for the preprocessors' statuses
1b2) the run will be submitted manually by Costin
1c) Recent runs with small statistics
1c1) detector experts should be careful when evaluating a period, since in case of negative returning codes (typical example is low statistics) they are not notified and CPass1 starts (since we use the validation codes now)
2) LHC12h
2a) Marian noticed 3 runs missing from his scripts:
2a1) 189699 --> failed in the snapshot (already reported at the beginning of November)
2a1a) Marian will add the check of the snapshot to his scripts
2a2) 192497, 511 --> to be understood
2a3) The runs won't be submitted (if it is found that it was needed), since LHC12h has already enough statistics
3) Merge-by-component
3a) Costin will resubmit the test with the modified executable
3b) Chiara will commit the changes in the executables to the trunk, as soon as she can commit (problems with svn)
3c) exit code: for the time being Costin will add that as soon as any aliroot process fails (even one partial merging), the whole process will be interrupted; the result should be equivalent to the current behavior
4) New TPC modules
4a) To be enabled, a change is needed in the AddTaskTPCcalib macro
4a1) Since Peter would like not to have any new revisions, we'll use it as input file, taking it from the trunk
4a1a) Marian will check with Mikolaj that it is ok
5) HMPID still out of CPass reconstruction
5a) this affects both CPass0 and CPass1
5b) Marian will add a comment to savannah asking Giacomo what the status is, since it would be important to enable the redo in CPass1 to have HMPID in the QA
6) TPC/Calib/Correction file used as input file in some CPass processed
6a) it will be simply taken as the other OCDB files from the snapshot; if problems come, we'll go back to the current solution (no specific tests could be done so far)
7) pA triggers
7a) in pA, no SPI will be used, so we have to find out the minimum number of events to have calibration for a run which lasted ~15 minutes (which is the criteria used now, to determine the SPI rate in pp), taking into account that:
7a1) The minimum number of SPI events that the TRD seems to need to have enough statistics is ~20000
7a2) The multiplicity in pA should be a factor 4 higher that in pp; we should anyway use factor 3, to be on the safe side
7b) in "min-bias" like mode (when all will be MB trigger) we'll downscale at redo level as we did for the three runs in LHC12c that did not have kCalibBarrel, and as was done last year; again, it is needed the minimum number of events
8) Wiping of CPass output
8a) proposal to remove all the chunk-level files, and keep only the merged one; the wiping procedure should occur after the PPass of the period under consideration
8b) Detectors are asked to prepare a statement for next week
9) LHC12h: calibration readiness for VPass
9a) Mean Vertex ready
9b) while checking the Mean Vertex it was found that the MeanVertexSPD object produced online is not ok for some runs
9b1) under investigation, also in terms of consequences for VPass
9c) TRD: waiting for QA feedback on 192707-708
9d) should be ready by the end of the week
9e) MUON alignment update will be postponed after VPass if everyone else is ready, to save time
9f) Some runs were with very little statistics (between 192585 and 192676): suggestion to mark them as bad
9g) Missing detectors contacted at the end of the meeting by Chiara
There are minutes attached to this event.
Show them.
