Andrew, Zukai, Gavin, Gennadiy, Jaroslav, Jan, Matthew, Jon, Yuri, Louise and Martin (taking notes)
- The DDT framework needs to be integrated with the DAQ messaging facilities.
- Andrew explains that we really need all of these DDS libraries in our framework so that we get the full messaging abilities.
- I brought up SNMP messaging and Gennadiy explains that SNMP does not guarantee delivery of messages whereas DDS does and hence needs more infrastructure.
- I will tell Gennadiy to put the message in source:trunk/DDTDemo/SendTriggerSignal_module.cc
- Gennadiy explains that he needs to rebuild libSim because it is not available for gcc 4.7.1. He will get this done by tomorrow.
- I will provide Gennadiy with a list of machines that we use and he will make sure that all of the products are set up on all of them. Here is the list: ** novadaq-ctrl-master.fnal.gov (NDOS DAQ Cluster) ** novadaq-far-master.fnal.gov (FD DAQ Cluster) ** novagpvm07.fnal.gov (NOvA VMs)
- Jon notes that there are several other things needed to get the messages sent properly. Andrew explains that the DDT framework will send messages directly to the GlobalTrigger until we get the DDT manager going.
- Yuri will start looking into the DDT manager next week after the Mu2e Collaboration Meeting is over.
- He changed the time slice time period to be more reasonable, but it might still need some tuning.
- He moved the calibration track fitter into its own package called "Tracking".
- The tricell trigger uses the tracks to then decide whether to make a trigger decision.
- He ran the code over 2000 events and triggers were successfully issued.
- I asked whether it is really necessary to run over high statistics MC using our trigger framework. Brian explains that he has seen differences between how his algorithm behaves in the Offline and DDT world because the slicing and tracking is different. Brian further notes that you generally want to be able to test your code in the framework you are using.
- This would be made easier by having individual MC files instead of 180 files. Jan seconds this motion.
- Andrew would like to test Brian's trigger on the FD cluster.
- I noted that the FD cluster does not have the DDT framework set up yet and that I am pretty busy now due to the Mu2e Collaboration Meeting. Gavin and I will take a look at the remaining setup and Gavin will finish it if it takes longer than an hour.
- I made a DDT tag for Jan using the Offline DAQ tag.
- After their conversation, Andrew reports that Peter would still prefer to make individual tags for Offline and DDT. No common tags yet!
- Andrew is asking for abstracts for CHEP in October in Amsterdam.
- Abstracts are due next Monday (4/1).
- Andrew proposes an abstract for a talk and several posters. Some possible topics:
- Zukai's pattern recognition algorithm.
- My MC integration between Offline/Online and DDT.
- My time profiling studies.
- Sorry that only my topics come to mind, feel free to add your own!
- Andrew will send an email to the DDT list requesting posters. He will include the abstracts from last year so we have a reference.