08/12/2014 Notes

DocDB Event


Pengfei, Kyle, Zukai, Alec, Matthew, Craig D., Yuri, and Martin (taking notes)

Andrew walked in as we walked out of the room, so we were able to tie up a few loose ends.

Online Monitoring

  • Kyle gave a talk showing his very nice progress on the DDT rate plots for OnMon.
  • He will consult with the OnMon experts to make a plan to get his code included into OnMon before he leaves, which will be presented at next week's meeting.
  • Kyle will use the long trigger name in the plot titles.
  • Kyle will make sure that his code of trigger bit extraction will not interfere while running over the numi and cosmics streams.
  • He will also check the normalization to see whether it is hard-coded to a 500 us or uses the actual event length.

Numu Trigger

  • I will set the auto-reconfigure to 1 and reset the prescale to 1.
  • I will take care of this when we roll in Zukai's updated fast monopole trigger.

Global Trigger

  • Both Alec and Andrew are actively working on it.
  • Alec confirms that the throttle works fine.
  • However, he noticed that the build is not very straightforward, so he will work on improving this.
  • Once the build is stable and easy, I will go ahead and test it on NDOS.
  • Activity Trigger Beam Veto:
    • We discussed the idea of implementing a veto on the activity trigger whenever a beam event is triggered.
    • Alec points out that this is slightly contrary to our data-taking philosophy of orthogonal data streams.
    • After some discussion, we came up with the idea of setting the NuMi bit in addition to the Activity bit when there is a beam event. This would allow for easy selection downstream.
    • Alex Radovich is interested in using this data, so Alec will start an email conversation with him to figure out what would be the best design.

DDT Streams

  • Zukai points out that the fast and slow monopole rates differ by an order of magnitude, so he would prefer to keep them in separate streams.
  • Matthew explains that one downside of small files is that more time is spent handling the files rather than running over it.
  • For DDT purposes, we would much prefer subrun merging into 1 file per run.
  • In the end, we decided to keep our 5 data streams the way they are: DDcalmu, DDnumu, DDslowmono, DDfastmono, DDenergy

Run File Stream

  • Alec suggests to look at the Datalogger XML configuration to look for the Config Block.
  • Andrew believes that this is somewhere deep in the Datalogger guts.
  • Andrew is happy with the way the data stream is currently figured out by extracting it from the file name.