Difference between revisions of "EPSCI SRO Meeting Jan. 11, 2021"

From epsciwiki
Jump to navigation Jump to search
Line 11: Line 11:
 
# ERSAP development
 
# ERSAP development
 
#* Indra VTP setup
 
#* Indra VTP setup
−
#** 1.2 GB/sec with 100% frame acceptance was achieved aggregating 2 VTP links, plus hit finding. Occasional hiccups were detected due to possible GC activities. David A. suggested to increase FE buffers. Chris L. offered to used his developed data-lake as a buffer between aggregator and hit-finding services. He will install his code base on Indra system: /daqfs/scratch/larreiu/src/sro-testsource
+
#** 1.2 GB/sec with 100% frame acceptance was achieved: aggregating 2 VTP links, plus hit finding. Occasional hiccups were detected due to the possible JVM GC activities. David A. suggested to increase FE buffers. Chris L. offered to used his developed data-lake as a buffer between aggregator and hit-finding services. He will install his code base on the Indra system: /daqfs/scratch/larreiu/src/sro-testsource
 
#* INDRA SAMPA setup
 
#* INDRA SAMPA setup
−
#**Ed J. led the discussion. The setup currently is equipped with ALICE FE card and ALICE express PCI C_RORC card, that can handle only 30Gbt/sec data throughput. Since FE electronics can provide up to 10GBT links , each 4.8Gbt/sec. TO avoid C_RORC bottleneck Ed will configure system to aggregate only 6 GBT links.
+
#**Ed J. led the discussion. The setup currently is equipped with ALICE FE card, and ALICE express PCIe C_RORC card, that can handle only 30Gbps data throughput. Since FE electronics can provide up to 10GBT links , each 4.8Gbps. To avoid C_RORC bottleneck Ed will configure system to aggregate only 6 GBT links.
−
#** ALICE C_RORC code was compiled and is ready for modification to add ERSAP components. In the raw ADC mode we can reconstruct pedestals and rare cosmic events. Unfortunately using of a radioactive source is forbidden in the INDRA lab. Within week our two the INDRA SAMP setup will be operational. We will use this time to understand LICE code and do necessary modifications.  
+
#** ALICE C_RORC code was compiled and is ready for modification to add ERSAP components. In the raw ADC mode we can reconstruct pedestals and rare cosmic events. Unfortunately using of a radioactive source is forbidden in the INDRA lab. Within a week our two the INDRA SAMP setup will be operational. We will use this time to understand ALICE code and perform the necessary modifications.  
 
# VTP Data Format
 
# VTP Data Format
−
# * David A. described steps required to implement VTP stream data format, based on EVIO-6 bank structure.
+
# * David A. described steps required to implement VTP stream data format (based on the EVIO-6 bank structure).
−
#** Ben will be implementing this data format on the Indra VTP setup. The question was raised weather aggregated stream bank (ASB) headers will increase data size. This might set some restrictions on aggregation/
+
#** Ben will be implementing this data format on the Indra VTP setup. The question was raised weather aggregated stream bank (ASB) headers will increase data size. This might set some restrictions on aggregation strategy.
 
# CLAS12 TriDAS & ERSAP  
 
# CLAS12 TriDAS & ERSAP  
 
#* Due to the time limitation this collaboration effort was not discussed
 
#* Due to the time limitation this collaboration effort was not discussed

Revision as of 21:11, 11 January 2021

Connection Info:

This meeting scheduled via MS Teams. Join link should be in your calendar.


Agenda:

  1. Previous Meeting
  2. Announcements
    • The next EIC SRO workshop scheduled for 17-20 or 28-30 April.
  3. ERSAP development
    • Indra VTP setup
      • 1.2 GB/sec with 100% frame acceptance was achieved: aggregating 2 VTP links, plus hit finding. Occasional hiccups were detected due to the possible JVM GC activities. David A. suggested to increase FE buffers. Chris L. offered to used his developed data-lake as a buffer between aggregator and hit-finding services. He will install his code base on the Indra system: /daqfs/scratch/larreiu/src/sro-testsource
    • INDRA SAMPA setup
      • Ed J. led the discussion. The setup currently is equipped with ALICE FE card, and ALICE express PCIe C_RORC card, that can handle only 30Gbps data throughput. Since FE electronics can provide up to 10GBT links , each 4.8Gbps. To avoid C_RORC bottleneck Ed will configure system to aggregate only 6 GBT links.
      • ALICE C_RORC code was compiled and is ready for modification to add ERSAP components. In the raw ADC mode we can reconstruct pedestals and rare cosmic events. Unfortunately using of a radioactive source is forbidden in the INDRA lab. Within a week our two the INDRA SAMP setup will be operational. We will use this time to understand ALICE code and perform the necessary modifications.
  4. VTP Data Format
  5. * David A. described steps required to implement VTP stream data format (based on the EVIO-6 bank structure).
      • Ben will be implementing this data format on the Indra VTP setup. The question was raised weather aggregated stream bank (ASB) headers will increase data size. This might set some restrictions on aggregation strategy.
  6. CLAS12 TriDAS & ERSAP
    • Due to the time limitation this collaboration effort was not discussed
  7. AOT

Useful References

Minutes:

TBD

Retrieved from "https://wiki.jlab.org/epsciwiki/index.php?title=EPSCI_SRO_Meeting_Jan._11,_2021&oldid=469"