EPSCI Group Meeting Aug. 10, 2020

From epsciwiki
Jump to navigation Jump to search

The meeting time is 10:00am.

Connection Info:

You can connect using BlueJeans Video conferencing (ID: 253 300 597). (Click "Expand" to the right for details -->):

Meeting URL
 https://bluejeans.com/253300597?src=join_info

Meeting ID
253 300 597

Want to dial in from a phone?

Dial one of the following numbers:
+1.888.240.2560 (US Toll Free)
(see all numbers - https://www.bluejeans.com/premium-numbers)

Enter the meeting ID and passcode followed by #

Connecting from a room system?
Dial: bjn.vc or 199.48.152.152 and enter your meeting ID & passcode

Agenda:

  1. Previous meeting
  2. Announcements
    • Beam delivery for physics ongoing (scheduled to end around Sep. 8)
      • David on shift 8/18, 8/19, 8/28, 8/29, 9/11, 9/12
      • Thomas on shift 8/20, 8/21, 8/30, 8/31, 9/9, 9/10
    • Streaming Data Scientist job posting (almost..?)
    • Fortnight Papers -> Move to separate meeting? Alternating Mondays with SRO?
    • NERSC User's Group Meeting 8/17/2020 @ 11am-6pm
  3. Graham's Project
  4. Ongoing Activities
  5. GUI for Calorimeter calibration scripts (Hall-D Request)
  6. Publications
  7. AOT



Minutes:

Attendees: David L.(chair), Carl T., Nathan B., Thomas B., Vardan G., Kishan R., Graham H.

Announcements

  • Fortnight paper discussions will be moved to a separate meeting bi-weekly on Mondays. These will be at the same time, but alternate with the bi-weekly SRO meetings.

JANA2

  • Most work over the past week has been related to A.I. support
  • Some communication with Ole on JANA2. (In consideration for SoLID)
  • A.I. support
    • Kishan was able to get python interpreter to work with JANA2 including event processor (no GPU inference yet)
    • Working with Nathan and testing interface (looking at calling interpreter from JFactory.)

A.I.

  • Kishan will try accessing the sciml190X nodes with the RTX titan GPUs this week
  • It was noted that he will need to generate a certificate (instructions on scicomp website)
  • epsci group does not have automatic access to scicomp systems like experimental halls do.
    • David will ask about this.

EVIO-6

  • Testing underway. Clearing seg. faults
  • Could use a second set of eyes with more C++ experience to review API and give comments
    • Nathan will do it once Carl sends info.

SPACK

  • Thomas has been communicating with Wouter who has put a lot of effort into implementing this for EIC
    • Working to minimize duplication
  • "ground is broken" as disk space and web access has been set up by (others in) the Computer Center

SRO

  • SRO helps minimize blind spots in physics data due to hardware triggers that must operate on limited information
  • Vardan is studying data lakes for low latency data movement
    • Single node prototyping being down at moment to investigate data performance.
    • Initial prototyping with Redis
      • Single threaded so multi-process must be used to get full speed(ignore earlier minutes reporting multi-threaded capability)
      • Bookkeeping becomes cumbersome it multi-process model is needed.
      • Original purpose for text data.
      • Tested 2 streams per redis process (12 streams total running simultaneously)
      • Will look at a few other possibilities (MongoDB seems popular)
    • Rapid prototyping progress shown last week at EIC SRO monthly meeting
    • Additional testing done last week
  • Ben R. has put together crate of fADC's in INDRA to use for SRO testing
    • Some question on cooling capacity of room since fADC's generate a lot of heat.
    • Possibility of moving some compute nodes to Computer Center room which has high cooling capacity.

OSG

  • 3 prongs of effort: oasis, xrootd, stash cache
    • oasis is in production and documentation has been given to Bryan.
  • Running OSG Jobs at JLab:
    • The head node has been stood up
    • No accurate timescale on when system will be open for jobs (main developers tied up with numerous other projects)
  • Primary OSG collector for JLab VO now running at JLab. Backup is UCSD
  • OSG category on ServiceNow has been added
  • Richard J. (UConn) has had some issues with OSG + Compute Canada
  • Graham: With broader use of OSG by multiple customers at JLab, we need a backup for Thomas in case he is unavailable
    • Thomas: Right now that would most likely be Bryan since he knows which developers are involved