This page provides details about the current implementation of the INT WFC pipeline for quick look reduction. It does not describe the pipeline reduction steps themslef but rather how the pipeline is used to get quick look data.

General Information

• Currently the pipeline runs on the INT data reduction computer. Once the Beowulf hardware arrives the pipeline will be transferred to it.
• We have run several pipeline processes in parallel on the INT data reduction computer and there seemed to be little impact on the computers responsiveness to the user (due to it being a dual processor computer with plenty of main memory).

What the observer should know

• The bias and flat frames applied to the data are the latest combined good calibration frames. Usually they are less than a week old. The directory for the quick look data (qldata$ under IRAF) containes the bias and flats used. Nameing of the flats is by the unique fiter ID - look for WFFID in the fits header.
• Quick look reduced data should be available about 5 minutes after a frame has been taken.
• To access the quick look reduced data through IRAF, prepend qldata$ to the file name and replacing "r" by "q". As an example, instead of
  cl> display r123456[1]
  use
  cl> display qldata$q123456[1]
  to display the quick look reduced frame.

Implementation

• LT 18:30 The quick look data directory is cleaned
• LT 19:00 - 08:00 cron runs a check for new files every minute
  • if another instance of the check is running exit
  • check observing data directories for new files
  • see if any of the files are target frames. If not, exit.
  • pre-process the target frame (takes about 1.5 minutes)
  • pipeline process the target frame (takes about 3.5 minutes)
  • move (and rename) the processed frame to the quick look data directory.
• LT 08:30 It is checked if all files were fetched and the files are then pre-processed.