OWG Meeting 28-Sep-2011

From GlueXWiki
Jump to: navigation, search

Agenda

  • Announcements
    • Nerses Gevorgyan should arrive 3-Oct
    • /group/halld-online
  • Review of minutes from 14-Sep-2011 meeting
  • Online agenda for GlueX collaboration meeting - Hovanes, Sasha
  • Networking - Elliott
    • hall
    • counting house
    • tagger building
    • infiniband in rack room
    • trailer?
  • Farm manager - Vardan G.
  • Review of other DAQ group activities for Hall D - Elliott
    • event disentangler
    • automated readout list
  • EVIO vsn 4
  • Controls
    • quench detector - Elliott
    • test stand - Hovanes
    • Solenoid Controls Redesign - Elliott
      • draft P&I, "final" in a few weeks
    • solenoid controls fast DAQ - Yi
  • CCDB online requirements - Elliott
    • configurations vs conditions vs calibrations
  • Elog - all
  • UPS power in the hall - all
  • PyIRMIS - Hovanes, Elliott
  • Multi-crate trigger test - Bryan
  • FDC cosmic ray test - Beni
  • Electronics, trigger and DAQ status reports - Chris, Alex, Dave, Fernando, Ben, Ed, William, Bryan
  • Mantis Task Tracker


Time/Location

1:30 PM Wed 28-Sep-2011 CC F326


Announcements

Next Meeting

1:30 PM Wed 12-Oct-2011 CC F326


Minutes

Present: Elliott W, Hovanes E, Dmitry R, Vardan G, Carl T, Simon T, Alex S, Mark I, Beni Z, Dave L.


/group/halld-online

Now exists with initial allocation of 10 GBytes. Accessible by people in the halldonline group (no "-" in group name, don't know why). Currently mounting this area must be approved on a case-by-case basis. Once the hall networks are set up access will be allowed via IP ranges.


Collaboration meeting schedule

The updated schedule includes talks by Alex and Hovanes. Due to increased activity it appears that one two-hour session at the collaboration meeting may not be enough. Later...I talked with Curtis and we will split the Online session into two sessions at the next meeting, most likely split being Electronics/Trigger and DAQ/Monitoring/Controls.


Networking

Plans for interim and permanent networking in the hall, counting house and tagger area have been worked out with the CC. Plans for getting interim networking to the trailer are still developing, most likely scenario is to run fiber to the trailer on poles.


Farm Manager

Vardan presented plans and options for development of a farm manager controls component. Note that this is a prototype for all controls components in the hall. Next step is to create a top-level Experiment Control Supervisor with an associated GUI. See his slides for more details.

  • Based on AFECS system used in DAQ run control.
  • Farm Control Supervisor (FCS) not necessarily a CODA component, but this is possible. Recommendation is to not do this.
  • Two agent types used: hardware control agent (HCA) and process control agent (PCA).
  • Both agents have multiple states, not necessarily identical to run control states.
  • One pair of agents deployed on each farm processing comnputer.
  • Both IPMI and SNMP supported, but IPMI is newer and has improved capabilities.
  • COOL database used to configure agents and specify states.
  • Some functionality overlap with NAGIOS and similar packages, can coexist without interference.
  • Many options for visualization: custom GUI, use EPICS BOY, etc.
  • Prototype code available now, report within a week or so.


Other DAQ group activities

  • Bryan is making progress on the automated readout list, a prototype will be used for the FDC cosmic ray test.
  • Dave A is working on the simulated raw event tape packer, should have a version out in a few weeks.
  • Carl is busy with the event builder/recorder (EMU) which we need for the FDC test this Fall. He has though about the event disentangler but not done much work on it.


EVIO vsn 4

Carl and Elliott plan to release a new version of EVIO this Fall, when Carl is done with the alpha release of CODA3. Major new features will include a built-in dictionary and unified file/stream/buffer I/O. Random access I/O may be included, depending on how difficult it is to implement.


Controls

  • A new quench detector may be needed, depening on results of Floyd's simulation of coil inductances later this Fall. The existing QD requires a very good inductance balance to operate well.
  • Hovanes continues working on the test stand, including controls for HV and LV crates, FCAL PMT's, etc.
  • See the Controls wiki for a full list of tasks for the group redesigning the solenoid control system.
  • Yi is testing a new National Instruments DAQ system that uses a single ADC per channel for the solenoid controls fast DAQ.


CCDB

Dmitry will soon work on implementing Mark's proposal for the Conditions database. Elliott made a clear distinction between reqirements for Configurations, Conditions and Calibrations.


ELog

Elliott noted that it is time to begin thinking about choosing an ELog for the hall. There are a substantial number of ELog-like applications and it would be nice to have a unified system.


Big UPS

The big counting house UPS is on order. Elliott asked everyone to let him know if they UPS power in the hall.


PyIrmis

Hovanes and Elliott reported that the IRMIS project has been superceded by a Python-based project, PyIrmis. First release expected this Fall. The basic schema remains but it is enhanced/extended in PyIrmis. Also, the web-based SOA interface is being replaced by a Python-based version with emphasis on bulk-manipulations from the command-line.