5/4/2015 Notes

DocDB Event


Alec, Jan, Leon, Craig G., Pengfei, Yuri, Andrew and Martin (taking notes)

Live Time

  • Pengfei included 74 buffer nodes last Wednesday and we ran with that until last Sunday without problems.
  • The chiller failed yesterday, so we went back to 45 buffer nodes.
  • Andrew says that we will move back to 74 buffer nodes with the run restart at noon today.
  • This is fantastic news and will keep us above 100% live time!

Up-Mu Trigger

  • Yuri added some preselection cuts to the slices and tracks that he does the tracking and merging on.
  • This should buy us some time.
  • Creating the associations between the tracks and the hit lists took quite some time because you have to loop over all of the hits.
  • Yuri now performs the preselection on the 3D tracks before asking for the associations.
  • Craig asks what the effect of the preselection is to the signal track selection.
    • Yuri will check this.
  • In the meantime, I can check whether the timing is improved.
  • Yuri points out that the main issue is that the finding of the associated hits is performed on all of the hits in the event instead of just the slice.
    • We will ask Brian to see whether it is possible to perform the association on the hits in the slice only.
    • This should be very fast then.


  • Yuri and I will be at test beam next week, so Andrew will run the meeting.