IPP Progress Report for the week 2010.09.06 - 2010.09.10

(Up to IPP Progress Reports)

Eugene Magnier

Serge Chastel

• Off on Tuesday
• IPP: not a lot this week (I worked on PSPS). I modified the script that displays active processes running on the cluster. I'll build up stats on Monday.
• PSPS: checking the contents of FrameMeta and ImageMeta works with one of Roy's data set (29152). I'm working on Detection. What to do next? SkinnyObject and ObjectCalColor in nominal case. I will then have to build up faulty samples to check if the tests are ok when something bad happens. I will also try to build up _tiny_ representative samples to speed up the tests. Roy's large original samples could be used in the future for performance tests (if it is required).

Heather Flewelling

Roy Henderson

• Vacation all week.

Bill Sweeney

• Examined the ThreePi.V1 dvo database. discovered that zero points were still not being applied. Also found that several exposures were added multiple times.
• fixed bugs in addstar application of zero points. Tested code.
• wrote a script to extract zero points from the gpc1 database and create a setphot input file.
• started setphot run on the ThreePi.V1 dvo database. This was interrupted by ipp004 disk problems.
• wrote script to parallellize rsync of ginormous dvo databases for backup.
• finished and integrated tools to simplify running of warp_skycell.pl and diff_skycell.pl from the command line for use in remaking corrupted files or running without threads
• debugged and fixed a problem with difftool -listrun that was breaking lookups done for detectability queries.
• Upgraded the postage stamp sever upload page to accept detectability queries.
• created pstamp.revert.dependent to reduce the amount of operator intervention needed for the postage stamp server.
• fixed problem with magicdstool -advancerun which caused failures to update the database to cascade.
• rebuilt 335 warpRuns so that the M31 warp variance images could be posted.(I accidentally cleaned them up before Johannes finished copying the files). Built the data store file sets by copying the files this time so that they will be unaffected by cleanup.

Chris Waters