[SRO] Meeting today

[Vardan Gyurjyan]

Dear All,

I'm sorry for the late notice. I just remembered this morning about a previous commitment to attend a play performance at HRA, where two of my granddaughters are participating. Therefore, I won't be able to lead today's SRO meeting.

Here are the detailed points from the agenda that I intended to discuss:

  1.  I'm working on a project to develop a user interface that allows designing, deploying, and configuring complete data-stream processing pipelines. This includes streaming readout and data acquisition capabilities, transport and distribution, and data-stream processing and visualization. The RTDP UI project will be a desktop application built using JAVA SWING, incorporating some design elements and programming solutions from the CODA JCEdit project. Additionally, we plan to develop a specialized RTDP ontology language based on RDFS, which will be used for the serialization, communication, and archiving configuration data.
  2.  Large-scale Deployment and Scaling Test of the CLAS12 Data-Stream Processing Application at NERSC through EJFAT LB at ESnet:
     *   Discussion on the outcomes from the recent 25-node horizontal scaling test.
     *   Challenges encountered in requesting additional resources due to the resource limits set for the EJFAT project.
     *   Jeng submitted a special resource reservation to reserve 40 nodes for 3 hours on May 20th at 10 AM. The submission was accepted this morning with the flag reservation=100g.
  3.  Ongoing Work to Debug and Commission the Return Path for Processed Events through the Second LB Back to JLAB:
     *   Readiness for local tests of the return path using two load balancer instances on the JLAB EJFAT testbed.
     *   Query on whether it is possible to request two local load balancers on the EJFAT testbed that are consistent with the latest LB upgrades and API.
     *   Procedure for releasing LB with the open end date ( -until 1970-01-01T00:00:00t00:00).
  4.  Developing a Demo for a Full Streaming Pipeline:
     *   Progress on developing a demo for a full streaming pipeline, starting from streaming readout of the FADC250 through the VTP, passing CTP aggregated time windows to CODA PAGG (primary aggregator) that will create two data stream connections: one for the critical path to ER (socket-based fast connection), and the second to the ET system that plays the role of the FIFO, presented data stream for ERSAP data-processing pipeline pulling data off the ET FIFO, identifying FADC hits, applying a simple trigger algorithm and visualizing triggered data in real-time.
     *   Request to update EMU to be able to accept multiple output channels with user-programmable prescale factors for two output channels, one critical channel that outputs all the stream events, while the second output channel outputs some percentage of the stream simultaneously.
     *   The current mechanism of the EMU provides multiple output channels, where each channel gets a stream unit in a round-robin fashion—seeking an elegant solution to avoid the stream unit copying required by the previous request.
  5.  Additionally, there are reports from RTDP, updates on the EJFAT project, and preparations for the abstract for CHEP24.

Sincerely,
-vardan


-------------------------------------------------
Vardan H. Gyurjyan, Ph.D.
Staff Scientist
Thomas Jefferson Accelerator Facility
Newport News, VA, 23606
E-mail: gurjyan at jlab.org<mailto:gurjyan at jlab.org>
757-269-5879 (JLAB)

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/sro/attachments/20240513/dd8b4703/attachment.html>