IPP Progress Report for the week 2010.03.01 - 2010.03.07

(Up to IPP Progress Reports)

Eugene Magnier

Heather Flewelling

Bill Giebink

  • Fixed bug in DNSM camera "heartbeat" software
  • Looked at PS2 cooling comunications error
  • Collected, reduced, and posted DIMM data
  • Weathered out doing RoboDIMM obs

Roy Henderson

  • Now removing duplicates, per exposure, before batching-up for PSPS. Slows-up code a fair bit. Hopefully a temporary hack.
  • Learned enough about MS SQL Server to clear out Dbs, query success/failure of loading etc.
  • Problem with missing chips in smf/DVO (0 detections): now skipping.
  • Merge into PSPS Db worked!
  • PSPS schema parser now produces XML table descriptions
  • 'init' batch values now parsed from XML to enable easy changes.

Paul Price

  • Reworked pswarp to prevent SEGV and/or failed asserts when failures occur in I/O.
  • Investigating warp, stack, diff failures: cleaning up error codes
  • Found bad warp files producing diff errors: o5246g0095o.137092.wrp.33343.skycell.2510.086.mask.fits, o5249g0523o.139274.wrp.34473.skycell.0943.104.mask.fits; recreated manually
  • Shepherding SAS stacks and z-band diffs
  • Fixed pmConfigConvertFilename so that file will be created even if Nebulous failed to create it.
  • Fixed PS::IPP::Config->file_exists to check that file is present on disk (not just in Nebulous)
  • Shepherding STS stacks and diffs
  • Fixed corrupt file: o5200g0229o.118727.ch.63604.XY11.ch.fits
  • Tracked down missing SweetSpot? data from Feb 22: convolved diff files had been removed which blocked magic. Not sure why this would happen (query checks magicRun.state!), but might be from a test run?
  • Added error-checking to source and PSF I/O
  • Queued 2010-02-11 SweetSpot?-Eve-South data
  • Found a bad copy of a raw file (consistently blocking chip processing): o5249g0334o.ota66.fits. Removed, leaving other copies.
  • Queued STS background-preserved processing and destreaking
  • Fixed stack zero points

Bill Sweeney

Chris Waters

  • Finished lossy compression code to free up disk space from engineering data and other images with minimal science value.
  • Burntool code in nightly science processing updated to correctly relaunch jobs that fail due to burntool or funpack failing. Also added blocking so that burntool gets priority over other computationally intensive tasks. This should allow burntool to finish quicker in the mornings, speeding up the nightly processing.
  • Scanned nebulous directories for orphaned and widowed files. Results posted at http://svn.pan-starrs.ifa.hawaii.edu/trac/ipp/wiki/Cluster_Storage_Notes. Started investigation to see if nebulous itself was creating the large number of new widow entries.
  • Preparation for UKIRT observing next week.