IPP Progress Report for the week 2010.08.30 - 2010.09.03

(Up to IPP Progress Reports)

Eugene Magnier

Serge Chastel

  • Czar (2 days): Wrote scripts to get active processes on the cluster nodes. Wrote wiki page for procedure when burntool fails.
  • Off on Friday.

Heather Flewelling

  • this is for 2 weeks, since I didn't do last weeks.
  • prepared a number of skybrightness and deteff plots for Ken
  • prepared for trip to Belfast
  • emailed Fabrizio about panstarrs disk usage (and consulted Jim and Gavin about this)
  • started/restarted ThreePi and MD dvo catalogs
  • started/stopped/investigated culling of raw images
  • answered many last minute ifaps1 liason questions before I left for Belfast.
  • attended Belfast conference

Roy Henderson

* PSPS

  • investigated loading errors from last week:
    • objIDs out of range: frames at bore-sight of -30 can have detections lower than this, but PSPS cut-off is -30
    • hundreds of frames failed to load: just a minor bug at PSPS end: fixed by Sue
    • continued loading 3PI all week, fixing minor issues as they arose

* IPP

  • czar on Monday
  • added more procedures to the processing wiki page
  • ippToPsps
    • Fixed bug whereby detections not included in final FITS file could count towards min and max object IDs for batch
    • new script to help when tracking source of load failures: for a given batch it will find the relevant smf as well as the corresponding ippToPsps generated FITS file for checking
    • added to cleanup script: now goes through and deletes files for batches that have reached 'merged' status
    • wrote a Batch class as part of my effort to clean up my Perl code
  • czartool
    • added logarithm plots to help when cleanup causes big jumps in processing data
    • some extended bug-fixing with the plots

Bill Sweeney

  • Spent 5 days as processing czar. (2 days assigned 3 days volunteer over holiday weekend).
  • Felt compelled to get rid of all outstanding faults. Wrote scripts to use gather data from database to regenerate corrupt files and to run failing commands outside of pantasks for debugging purposes. As of 5am Sunday morning there were no outstanding faults in the enabled labels. Let's keep it that way, please.
  • At Ken's request made some postage stamps for George Herbig of some interesting stars that have been reported to have brightened recently. It was an interesting exercise. It's remarkable how often specific things that we want to look at are masked.
  • Started regenerating the warp variance files for MPG. It turned out that the data store was cleaned up prematurely. Unfortunately the underlying data was cleaned up as well.
  • updated the postage stamp web interface to use a program that supports the full postage stamp server request specification. Added bycoord lookups but didn't announce.
  • Created and started a program to delete diff images that were missed by destreak cleanup. It noticeably adds to the load on our nebulous data base so it is running serially. It should recover many TB of data over the next couple of weeks.

Chris Waters