Home » Groups » STAR Protected

Differences between SL09g and SL10j for run9 pp500 - IV

Updated on Mon, 2011-03-07 09:30. Originally created by seelej on 2011-03-02 09:45.

Emboldened by the information that the software bug between SL10f and SL10g is not the only problem causing the lost tracks between SL09g and SL10j (blog) and that some entry in the DB seems to be the source of the difference I began to search DB timestamps for when the decrease occurred. By running small (300 event) productions I have narrowed it down to a single entry time day, 2010-01-29. 

 

I then searched some relevant DBs (Calibrations_tpc, Calibrations_tracks, Calibrations_rich, Calibrations_rhic, Geomtry_tpc) to see which, if any, of them had new, relevant, entries on that day and indeed two of them do, Calibrations_tpc and Calibrations_tracks. The tables in those DBs that changed are

 

Calibrations_tpc 
 
TpcAdcCorrectionB
TpcDriftDistOxygen
TpcLengthCorrectionB
tpcPressureB
TpcSecRowB
TpcZCorrectionB
TpcZDC

 

 
Calibrations_tracker
 
svtHitError
tpcInnerHitError
tpcOuterHitError
tpcTrackingParameters
 
 
After doing the DB difference search I then tried to find exactly which DB entry was the cause. I found that the DB change that caused the difference in the number of reconstructed high pT tracks was a change to the tpcInnerHitError and tpcOuterHitError tables in the Calibrations_tracks DB. I am not sure what the parameters in the DB are, but they all became much smaller than they had been previously with this DB entry. 

 

 

Groups: