IPP Progress Report for the week 2011.12.12 - 2011.12.16

(Up to IPP Progress Reports)

Eugene Magnier

Serge Chastel

Heather Flewelling

• czar/summitcopy investigator
• ifaps1 liason
  • answered questions for Klaus and Tom
  • transferred a copy of LAP dvodb to ipp022
• investigated psf_major changes between tags (yes, there are changes, yes, the current tag is the correct calc)
• further rsync testing of the big dvodb. It's slow (4-6 hours between minidvodb merges)
• addstar work continuing (for better ways to handle merges of minidvodbs)

Roy Henderson

• PSVO:
  • added sub-menus to XML plugin query files
  • wrote recursive function to parse sub-menus
  • plugin files now online so PSVO clients can update dynamically (defaults to local copy if offline)
  • added more queries
  • released new version

• Loading:
  • usual loading monitoring/czaring
  • processed some of the stragglers from the OLD 3pi processing (some were missing from Gene's backup DVO I had been using)
  • yet more dupes during merge. Worked on a list of batches for Sue to remove.

• ippToPsps development:
  • datastore delete failing. Spent ages tracking it down. My mistake: fixed.
  • working on clients creating their own scratch Dbs using PID as part of name to avoid conflict and ease of client-start-up

• looked into providing plots for Heather to asses load on ipp005 during loading

Mark Huber

• finished stepping through ppStack and working on summary tests.

Bill Sweeney

• Spent this week implementing the new image processing pipeline stage static sky calibration.
  • new staticskytool modes
  • new script
  • tweaks to recipes and file rules for psastro
  • code to read in Extended source parameters so that they can be saved.
• Changed the postage stamp server and update processing allow postage stamps to be released without magic censoring. Horray!
• Changed distribution to allow nightly science data to be released uncesored.
• Cleaned up distribution mess caused by removing the magic requirement. Old data got queued for distribution when it was no

longer available.

• Reran M31 raw stage distribution to replace filesets that were damaged by the loss of the ipp064 RAID

Chris Waters

• Finished testing pattern correction update/background changes. Although the background sigmas and numbers of detected objects do not seem to change much, the jpeg images look improved in the new version.
• Fixed bug in shuffle code that could allow two instances of a file to exist on one host. Restarted shuffle.
• Mailed analysis of diskspace results to Gene and Ken.
• Implemented changes to SweetSpot nightly science reduction to allow stacks to be made. Have not enabled WSdiffs for this because of a race condition between diff queuing and stack completion.