S&C plan, PWGC 201406


First off, please note that a brief summary was provide at the S&C meeting two weeks ago and available at You do not have access to view this node.
The same summary was also given to Frank, the management team and the local BNL group. This is "only" the 5th time ... I would appreciate for the PWGC to pay more attention to S&C topics.


The status of production more or less left past production incompletely done or with problems. As we remember
  • Run 12 data
    (p+p, U+U, Cu+Au) was completely done but has showed or had issues
    • The Cu+Au data samples were done with best attempt calibration SC&GL method has been a questionable item for years for asymmetric system as it relies on matching East/West symmetry.
      Status: A new method R&D has been thought - a service task is ongoing on this (more later)
    • Trigger issue with VPDMB - code fixed but only st_W data reproduced. There is no more to do with this than reproduce the data.
    • The h+/h-  issue reported in 2012/08 analysis meeting is still there but partly hidden
      1. the executive decision at the time to revert to the Year 2000 alignment has not resolved the "why" but  decreased the effect
      2. There is a clear issue not well understood and ...
      3. The most central bin (0-5%) of U+U indicates the split is still there.
    • h+/h- note: This issue was seen since Year 10 data (some hints in 2008 but was not clear) ...
      An activitty to revisit the TPC static distortions have been ongoing for a while.
  • Conclusion: Run 12 would highly benefit from a full re-production to solve alignment trigger and possible asymmetric systems
    Reproduction - 2 months max, would be done at KISTI of resources are still available (that was our initial plan, news from KISTI tend to indicate they will not renew their MoU)  but Grid based framework is ready.
     
  • Run 13
    We skipped days 129-161 having the PXL in due to an issue of Geometry implementation. We could see the jump between the two period through the BeamLine calibration as follows ...

    We only produced ~ 1/2 of the dataset pending PXL geometry without which, we CANNOT reliably track in Run 13 OR Run 14 data.
    TOF, dE/dx and TPC hit errors needed for the data ... and of course, beamLine calibrations.
    Conclusion: 2 B events ~ 5-6 months with current BNL resources for the full production but 1/2 that for the second part. Shooting to produce the second part consistent with the first 1/2 calibration method (same method and accuracy) and reserve a second pass production possibility whenever all calibration / pending issues are resolved.
     
  • Run 14
    A sample of the BES was done but with approximate tracking, problems with the PXL geometry and a few other issues (T0 was wrong) - we KNOW this data preview is NOT to final result quality due to those many problems.
    Other: Production timing done based on TPC only chains so far - we have some unknown in delivering the final numbers including (a) including all HFT releated code in the chain and (b) knowing the outcome of the procurement cycle aimed at buying new resources (cost of hardware has increased / not sure we can continue to ride the curve of resource increase).
     
  • AND ... We have many pending tasks
    • Code reviews of the SST, MTD
    • Incoming HCAL with no code we know off
    • Re-assessing Stv progress and move on (we are dangerously on a plan B mode)
    • Need to move forward with the VMC project
    • Need to move on with a newer ROOT 5x version ...
    • ... will pass on infrastructure reshapes ...

Bottom line
  • Our plate is full - we are NOT opened for a fast production of (for example) He3+Au which will distract people away from critical tasks ... least speaking of the BES production generating community discussions based on a production we know has problems.
     
  • The Geometry for the PXL is a dire show-stopper and has to have priority
    Without this, we cannot move on neither Run 13 nor Run 14.
    Status: This activity is directly coordinated by S&C now on a bi-weekly meeting basis. Progress are reasonable, thanks t lots of work from Dmitri Smirnov ... but we probably need more time (a status report / checkpoint will be made next week).
    If successful, we would produce the second part of Run 13 as a test and perhaps use the BeamLine test (which showed a divergence between intercept between the two period).
    This is self contained and an internal check - no help neded.
    Before rushing to Run 14 data, we will need the other task to come inline to save time ...
    Run 14 production is projected to be lengthy - we wil know the exact time when we will know (a) the speed impact of having all code from the HFT folks in and (b) what resource we have / can purchase on top of what we have now ...
     
  • The second priority is the static TPC distortions. This work is being handled by Yuri and progress
    Status: status (also coordinated in a bi-weekly meeting) as per yesterday is encouraging.
    The data sample to be used to verify this will be Run 12 data.
    Run 12 is chose to avoid dealing with the HFT geometry so both tasks can go in  //.
    Note: If successful, Run 12 will be reproduced AND the PWG will be asked  to look into the samples (U+U included) and verify physics outcome. Ideally, previous production would be deleted and replaced. 
     
  • The SC&GL method for asymmetric system study and R&D is ongoing. 
    A student will work on this in July end-of-August but has already started (during a visit for shifts).
    The base sample will be Cu+Au - we will reproduce this and see.
    This task will be independent of everything else (also a strategy to maximize the task and problem solving parallelization). However, if the static alignment pans out, we will also need to re-run, as a final check, the new SC&GL before producing a new Cu+Au sample.
    Note: Same, PWG will be asked to look into this.

Again, we have our hand full and would unlikely be in a position to produce samples before early September.