2016 September 30th
|Tip: The PDF version of this document is more suitable for printing.|
This documents a general plan for data acquisition at the Canada-France-Hawaii Telescope (CFHT). It attempts to cover most relevant software development practices and computing hardware directly involved in the process of controlling a science instrument, reading pixels from a detector (infrared or visible), packaging the pixels as data files, and visualizing the image. The target goal is to accommodate wide field imaging devices like MegaCam and WIRCAM, and provide a general framework for future instruments at CFHT. As of 2003, most of the important infrastructure for status, logging, and command processing is now complete.
Our current computing architecture and goals for the future are the result of careful consideration to available hardware solutions, compatible operating systems and software, and a path of transition from our existing infrastructure (which we came upon by the same methods in the past.) The evaluation, planning, and implementation of these goals are iterative processes, so the things presented here may change as technology improves. In summary, this document presents a general philosophy, the current status (last updated 2016), and our best prediction for the near future.
While other software systems at CFHT are at least partly based on the architecture set forth here, there are many which are not described in this document. They include the Telescope Control System, the Elixir project which provides sophisticated data analysis, the Queue Scheduling project which plays the role of an efficient observer during survey mode and Queue Scheduled programs, and the data archive pipeline which packages data for distribution and transfer to other facilities. The information here is limited to what is most relevant to instrument designers and acquisition interface designers. The users of these systems will want to read the observing manuals instead. This document is for developers.