Project

General

Profile

04/15/2014 Notes

DocDB Event

Attendance

Alec, Craig D., Matthew, Andrew, Zukai, and Martin (taking notes)

General

  • EC decided the following:
    • turn the cosmic pulser down from 40 Hz to 10 Hz
    • turn up TriCell trigger
  • We would like to turn TriCell up by a factor of 8:
    • factor of 4 to compensate for pulser turn down
    • factor of 2 to compensate for doubled detector size
  • This means that the prescale will be changed from 1000 to 125, but we'll first consult with Brian about this.

Live Time

  • Andrew suggests to probe occasionally and report.
  • Craig suggests to run the trigger over the pulser data, make sure that all gets caught and then determine an efficiency.
  • The BNEVB could report the total_missed number to Ganglia, but Andrew notes that this is not a good record due to Ganglia's sparsification.
  • In the end, we decided to write a program to populate a log file as follows:
    • time stamp, number of missed milliblocks
    • if accessible, we can also add the run and subrun number
  • I will write this program.

Data Stream

  • The production group wants us to combine our data streams to reduce the number of files.
  • Andrew suggests that we group by rate as follows:
    • DDcalmu (3 Hz going to 24 Hz)
    • DDnumu (14 Hz)
    • DDexotics = DDenergy + DDfastmono + DDslowmono (a few Hz)
  • We will coordinate with Louise to get this implemented.

Data Overlap

  • All data streams are independent, so you do have to run over both streams to find overlap.
  • Andrew notes that this was a compromise made when designing the DAQ system.

Nearline

  • We discussed that a closure test would be useful in the Nearline.
  • Basically, we want to make sure that the trigger rate based on the pulser data is the same as the raw trigger rate.

FD 7 DB Streamer Data

  • I will request a streamer data run at the Outfitting meeting tomorrow, so we can tune our triggers.

FEB Overflow

  • We will put this trigger into a new package: DAQStatusTriggers.
  • Andrew notes that we have to set the trigger up to ignore FEBs that have already been seen with the overflow error bit because this bit latches.
  • Before improving the trigger though, let's first see whether it will even trigger.

DAQ Experts

  • Matthew will not be able to do DAQ expert shifts.