IPP Progress Report for the week 2011.03.07 - 2011.03.11

(Up to IPP Progress Reports)

Eugene Magnier

Serge Chastel

• Updated Database Backups Scheme
• Created IPP Databases
• Created Apache Tuning

Heather Flewelling

• sick half of Monday and half of thursday
• ippconfig iterations (many) - finally is working - need to investigate camera stage to look for mask tweaks.
• addstar - dvoverify can check if sorted (-s), and can tell the first unsorted file it encounters (-s -v), started working on (designing) changes for addstar to be more intelligent about verifying databases.
• created gpc1gcirc.pl - finds the nearest raw 'sciency' (o%g%o) exposures for a given ra and dec. located in tools
• czar friday - found raw files with inconsistent copies:
  • gpc1/20090628/o5010g0081o/o5010g0081o.ota45.fits
  • gpc1/20090628/o5010g0082o/o5010g0082o.ota24.fits
  • gpc1/20090628/o5010g0083o/o5010g0083o.ota45.fits
  • gpc1/20090628/o5010g0083o/o5010g0083o.ota52.fits
  • gpc1/20090628/o5010g0085o/o5010g0085o.ota25.fits

Roy Henderson

• PSPS
  • retrieved more MD04 files for Jim
  • loaded more batches, including an init batch, for Conrad to test the DXLayer
  • wrote a class to encapsulate terrible cfitsio library. Cleaner, safer.
  • worked on parsing model fit table from Gene's newest stack cmf file
  • improvements to unit test for ippToPsps
  • figured out how to make a nice Mac OS app version of PSVO

• IPP
  • Czar on Monday. No data.
  • added to ippToPsps and czartool documentation on wiki
  • started looking into porting Heather's nice ra/dec image locater script into PHP for ippMonitor

• Other
  • Sick on Tuesday
  • Doctor's appointment Wednesday morning

Bill Sweeney

• Some minor postage stamp server wrangling. It has gotten much better at taking care of itself.
• Modified diff_skycell.pl and stack_skycell.pl to replicate precious output files. Tested changes and integrated into the production branch.
• Modified the distribution system to facilitate the distribution of raw detrend images. Distributed raw detrend images for MPG.
• Continued working on analysis of the sts astrometry problem.

Chris Waters

• Reorganized ipphosts.mhpcc.config to balance the disk load across the cluster. This should utilize hosts with larger disks more than hosts with smaller disks.
• Finished diffraction spike updates. New wider masks should cover the majority of spikes correctly, except where proper motion has moved the star away from the expected position.
• Wrote script to identify and truncate large (> 30MB) logfiles. These usually result from jobs that fail repeatedly, so the truncation should not eliminate useful information. Needs to be parallelized and run across the cluster.
• Various tests to attempt to determine a good initial reprocessing test set.
• Bugfixes to ppImage to handle format/recipe setting of PATTERN correction correctly.
• Create updated DQstats detection efficiency CDFs. Incomplete, due to a lack of z and y band images.
• Checked the timer values created at each stage and compared to database. Added dtime_convolve to ppStack, as this takes up the majority of the processing time. Need to test database change and code changes before committing to trunk.
• Identified slow processing of chip XY06 as being due to electronic noise on some cells. Created mask to block these cells while camera team worked on resolving the issue. After fix, discovered that there are still single pixel rows that have the electronic noise still.
• Added options to dettool -updatedetrun to allow the start and end times of a detrend to be set via the tool.
• Re-read detectability server code and identified a number of bugs that are probably the source of the trouble.