IPP Progress Report for the week 2012-07-30 - 2012-08-05

(Up to IPP Progress Reports)

Eugene Magnier

Serge Chastel

  • Enlarged storage nodes disk view in czartool
  • Fixed a few lost otas
  • Updated ppMops for trails fitting
  • ippc63 replicates ippdb00 nebulous database

Heather Flewelling

  • czar - registration problems on monday - I fixed them manually, chris committed fixes to fix them more permanently
  • ipptosps
    • schema changes for new improved columns in objects
    • changes from staticsky to skycal
    • note to self: with new tag and new dvo schema, kill all screen sessions and recompile dvograbber or else you get junk in batches.
    • friday afternoon: sa8 is loading with new dvo schema, skycal, and new ipptopsps schema.
  • dvo
    • skycal is implemented in addstar and committed in trunk
      • to be used instead of staticsky
    • Sa8
      • first attempt to build - failed with fault 2 and I don't know why
      • second attempt worked
      • relphoted with special gene command
    • ThreePiQ2
      • failed, two problems:
        • 2 minidvodbs are not quite right - they pass dvoverify but cause lots of messages by dvomerge
        • perl script grows too big while holding onto the messages caused by dvomerge (gene found this)
      • rebuilding, without the 2 funky dvodbs
    • skytable.fits - whatever problem it was, it's not repeatable. Going to ignore it for now...
  • psps
    • investigated on the sandbox objects - ken had found funny values in the columns - I expect this is due to the swaps of column names, the batches themselves look good.
    • made new sa8 test batches for psps sandbox
  • science stuff for durham

Mark Huber

  • MD:
    • MD01 refstack moved to high priority with nightly science data being taken. Handful of skycells being problematic with incorrect scaling and rejection.
    • MD10 refstack rerun attempt when weight map arc of hight values found and make more homogeneous, still found some problematic skycells.
    • MD04,09 deepstack test to flush out more full field processing bugs and provide another deepstack comparison sample to approximate 2yrs (MD04 older Nov. 2011 processing and MD09 newer June 2012 processing).
    • MD staticsky/skycal rerun (with SAS_v8) after deteff extension added and INST_FLUX column poplulated for dvo/PSPS.
    • MD.GR0 -- pulling together bits for posters and talk, initial depth estimates for Bob Joseph.
  • Czar Friday: summitcopy+registration has been temperamental, some mount issues, psastro WCS found off solution for mis-pointed 3PI exposures.
    • looked into odd diffim magnitudes for sample from Peter.

Bill Sweeney

  • Continued working on psphotStack. This program takes a long time to run and thus debugging is slow going.
  • Found and fixed the cause of the memory corruption problem that we worked on last week. Turned out that the code was using a non-thread safe data structure.
  • Found and fixed a race condition that caused a SEGV in the extended source fits.
  • Modified psphotStack to enable Detection Efficiency measurements.
  • Started an exhaustive analysis of the memory usage in psphotStack. With all of the bug fixes and optimizations introduced over the past several weeks it now completes in reasonable time (1-2 hours) for a skycell near the galactic plane, but it still grows to over 32 GB in virtual memory usage in those fields.
  • fixed a postage stamp problem that I inadvertently introduced last week.
  • updated the list of preferred data_groups for postage stamp requests. With the switch to the ipp-20120802 tag that code is now deployed.
  • found and fixed cause of an assertion failure in the extended source fits.
  • 2 days as processing czar.

Chris Waters