Difference between revisions of "SRO RTDP Sep. 1, 2023"

From epsciwiki
Jump to navigation Jump to search
 
Line 9: Line 9:
 
** Computer (expected Sept. 5)
 
** Computer (expected Sept. 5)
 
** 2x U280 FPGA cards (delayed at least 3 months from computer order)
 
** 2x U280 FPGA cards (delayed at least 3 months from computer order)
** Network Switch (expected in October)
+
** Network Switch (expected in October). (Marco offered to look for temporary loaner he could ship to JLab if the ordered one is delayed.)
 
* The switch is intended to host fibers from all VTPs. The switch will reside in the CH.
 
* The switch is intended to host fibers from all VTPs. The switch will reside in the CH.
 
** Fibers are not yet installed to connect all VTPs. Only about a dozen are available.
 
** Fibers are not yet installed to connect all VTPs. Only about a dozen are available.
** Sergey just received word from Chris C. the day before that the installation will not happen until next summer. No further details available yet.
+
** Sergey just received word from Chris C. the day before that the fiber installation will not happen until next summer. No further details available yet.
 
* VTP Modules:
 
* VTP Modules:
 
** The forward carriage has 18 VTPs with 6 more ordered.
 
** The forward carriage has 18 VTPs with 6 more ordered.
Line 20: Line 20:
 
** Some CAEN TDCs will be replaced by JLab TDCs next year.
 
** Some CAEN TDCs will be replaced by JLab TDCs next year.
 
** There are 3 other board types that are not ready for streaming. These need firmware that Ben Raydo is working on. He is overbooked and the timeline on those updates is not known.
 
** There are 3 other board types that are not ready for streaming. These need firmware that Ben Raydo is working on. He is overbooked and the timeline on those updates is not known.
 +
*** David will talk to Ben and FEE to learn what the expected schedule is
 
* Hall-B storage includes a RAID disk capable of 2-3GB/s
 
* Hall-B storage includes a RAID disk capable of 2-3GB/s
 
* Patrick Noted that Run Group E will also use a LH<sub>D</sub> target in addition to the nuclear targets
 
* Patrick Noted that Run Group E will also use a LH<sub>D</sub> target in addition to the nuclear targets
 
** David will update the test plan.
 
** David will update the test plan.
 +
* There was some question about the bandwidth upgrade from the CH to the CC. David said he would follow up with Brent on that.
 +
 +
We discussed variations on the testing plans involving Hall-B. These included:
 +
* Take data from portion of CLAS12 using the existing fibers
 +
* Purchase short fibers that could be installed ourselves and place the new switch in the hall temporarily
 +
Sergey will discuss the original installation plan with Chris to understand why it is being delayed so long.
 +
 +
 +
We also discussed the nature of the High Bandwidth tests. Two scenarios were discussed:
 +
* Stress the bandwidth of the components we have by triggering on noise
 +
* Read out as much of CLAS12 as possible at as high of bandwidth as possible and capture to files
 +
 +
 +
We agreed that exact plans will depend on actual schedules of: delivery of switch, fiber installation, and firmware upgrades
 +
 +
<hr>
 +
Action Items:
 +
# Send network switch model to Marco so he can try and locate a backup
 +
# Find out details on delay of fiber installation in Hall-B (Sergey)
 +
# Find out schedule for upgrading CH to CC network bandwidth (David)
 +
# Find out VTP development schedule for supporting additional module types in Hall-B (David)

Latest revision as of 14:42, 1 September 2023

Attendees: David L., Patrick A., Sergey B., Mariangela B., Vardan G., Marco B., Markus D.


Minutes

  • We referenced v0.0 of the Test Plan document during the discussion
  • Hall-B currently has 3 relevant items awaiting delivery
    • Computer (expected Sept. 5)
    • 2x U280 FPGA cards (delayed at least 3 months from computer order)
    • Network Switch (expected in October). (Marco offered to look for temporary loaner he could ship to JLab if the ordered one is delayed.)
  • The switch is intended to host fibers from all VTPs. The switch will reside in the CH.
    • Fibers are not yet installed to connect all VTPs. Only about a dozen are available.
    • Sergey just received word from Chris C. the day before that the fiber installation will not happen until next summer. No further details available yet.
  • VTP Modules:
    • The forward carriage has 18 VTPs with 6 more ordered.
    • Each VTP has 2x10Gbps links
  • Other Modules:
    • Currently, only the FADC modules are capable of streaming.
    • Some CAEN TDCs will be replaced by JLab TDCs next year.
    • There are 3 other board types that are not ready for streaming. These need firmware that Ben Raydo is working on. He is overbooked and the timeline on those updates is not known.
      • David will talk to Ben and FEE to learn what the expected schedule is
  • Hall-B storage includes a RAID disk capable of 2-3GB/s
  • Patrick Noted that Run Group E will also use a LHD target in addition to the nuclear targets
    • David will update the test plan.
  • There was some question about the bandwidth upgrade from the CH to the CC. David said he would follow up with Brent on that.

We discussed variations on the testing plans involving Hall-B. These included:

  • Take data from portion of CLAS12 using the existing fibers
  • Purchase short fibers that could be installed ourselves and place the new switch in the hall temporarily

Sergey will discuss the original installation plan with Chris to understand why it is being delayed so long.


We also discussed the nature of the High Bandwidth tests. Two scenarios were discussed:

  • Stress the bandwidth of the components we have by triggering on noise
  • Read out as much of CLAS12 as possible at as high of bandwidth as possible and capture to files


We agreed that exact plans will depend on actual schedules of: delivery of switch, fiber installation, and firmware upgrades


Action Items:

  1. Send network switch model to Marco so he can try and locate a backup
  2. Find out details on delay of fiber installation in Hall-B (Sergey)
  3. Find out schedule for upgrading CH to CC network bandwidth (David)
  4. Find out VTP development schedule for supporting additional module types in Hall-B (David)