OWG Meeting 10-Apr-2013

Agenda

• Announcements
  • Still need to arrange for Ed to talk more on FPGA programming
• Review of minutes from 27-Mar-2013 meeting
• Status and schedule of Hall D projects proposed to be done by DAQ group - Graham
  • Linux SBC configuration, installation, booting - by Bryan
  • InfiniBand hardware specification, purchase, installation - by Dave A
  • Farm manager CODA component - by Vardan
• Implementing "deferral" in L3 farm - Graham
• Online directories and file structure - Elliott
• Compiler and interpreter versions - Elliott
• Status of DAQ testing in Rack Room - Dave A, Bryan
• Counting House update - Hovanes, Elliott
• Solenoid test update - Elliott
• EPICS update - Hovanes
• Controls update - Elliott, Hovanes
• JInventory database - Elliott
• DAQ group projects- Bryan, Dave A, Graham, Carl, Vardan, Elliott
  • EVIO, cMsg, CODA3, ECS/CSS, disentangler, readout list, Java/cMsg GUI
• Electronics, trigger and DAQ status reports - Chris, Alex, Dave, Fernando, Ben, Ed, William, Bryan
• Mantis Task Tracker
• Midas Elog

• Upcoming topics
  • Calibration/Alignment issues?
  • Status of processing uncompressed and compressed simulated raw data and monitoring/L3 farm - Dave L, Elliott
  • What online projects can other groups take over? - Elliott

Time/Location

1:30 PM Wed 27-Mar-2013 CC F326

Next Meeting

1:30 PM Wed 10-Apr-2013 CC F326

Minutes

Present: Elliott W, Graham H, Dave A, Bryan M, Hovanes E, Dave L, Simon T, Sean D, Beni Z, Mark I.

Deferral in DAQ System

Graham presented various deferral schemes we could implement in the DAQ system. See his slides for many details.

• Ours is he first full L3 farm at JLab.
• Many farm computers must be running to do L3 online, w/o deferral.
• L3 typically has a different computing processing model than offline or DAQ computers.
• Deferral is in many ways simpler
• Can use file system on L3 computers or on separate system to buffer events.
• L3 processes can get events from files, simplifies end-or-run.
• Many tradeoffs between the four different models presented.\
• Questions concern event order, file concatenation, role of farm manager, disk space needed, etc.
• All options require writing to disk at 3 GB/s.
• No extra CODA development needed for any option, already can write to disk files.
• Complications appear in farm mgr and L3 node software.

Directories and Compilers

Elliott presented proposals for directory structures, accounts and compilers. See link above for many details. Plan is to use full C++11 standard for all new code.

DAQ Testing

Dave A and Bryan M presented the status of DAQ testing in the counting house.

• 7 crates installed including TS.
• Getting cables, may have to move a crate.
• Trigger distribution should be done by the end of the week.
• ROCs booting properly, still need to collect a few cpu's.

Counting House

Hovanes reported that gluon20-23 are installed and working. The PR for seven servers is in.

Solenoid Test

Elliott reported that the cold box still has carbon contamination problems, but we should be cold and full of LHe soon. Many problems with the Danfysik supply, but it is ready to go.

EPICS

Hovanes reported:

• Running on gluon20.
• Have to move from 36 to 194 subnet.
• Need to install some HV/LV crates in counting house.
• Beni continues to work on HV/LV gui's using info from Fernando.
  • where to store config history? JInventory? conDB? SVN? other?

SCONS

Elliott will meet with Maurizio and learn how to use his system. Joint development with CLAS12 is possible.