IPP Progress Report for the week 2011.04.04 - 2011.04.08

(Up to IPP Progress Reports)

Eugene Magnier

Serge Chastel

• Databases backup. All details in the wiki ​http://svn.pan-starrs.ifa.hawaii.edu/trac/ipp/wiki/DatabaseBackups
• Mops Hemocrat
• bzip is better than gzip to compress dump files (and possibly products?)

Heather Flewelling

Roy Henderson

• Some design, discussion and a little coding for the implementation of Daniel's query builder into PSVO. Should be ready for Boston....

• Lots of development on ippToPsps, culminating in the loading to PSPS of the first-ever stack format batch, albeit with bogus object IDs. The road to this included:
  • now using local, rather than remote, MySQL database to enable a huge speed increase when loading tables from FITS
  • wrote a method to report which columns of which tables are fully or partially NULL, and to replace those NULLS with weird PSPS '-999' pseudo-NULL value
  • many more stack fields populated after discussions with Jim/Gene
  • implemented XML config file for use with Jython code: includes all Db parameters, paths, DVO and datastore settings etc
  • wrote a python class to handle gpc1 queries, including nebulous calls to return file paths
  • 'DVO' table in database to store all IDs (objID etc). Filled with fake values and used to test load
  • jython code now interfacing with ippToPsps db, getting new batch ID, saving manifest files and publishing to datastore
  • worked with Thomas and Sue loading the stack to the ODM. A few problems encountered, but it loaded eventually
  • started work on Jython version of detection batches, though this is low priority

• Meetings and discussions about IPP processing required before Boston meeting to ensure a good PSPS demo.

• Worked through a list of exposures missing from PSPS as reported by Dave Monet. Figured out a good reason for most, if not all.

Bill Sweeney

Chris Waters