The URLs that appear in test_one recipe example (Sec. Test the installation), reference web services available within the Gemini Observatory’s operational environment. They will not be available directly to users running reduce outside of the Gemini Observatory environment.
In the context of reduce and the Astrodata Recipe System, FitsStorage provides a calibration management and association feature. Essentially, given a science frame (or any frame that requires calibration) and a calibration type requested, FitsStorage is able to automatically choose the best available calibration of the required type to apply to the science frame. The Recipe System uses a machine-oriented calibration manager interface in order to select calibration frames to apply as part of pipeline processing.
Though this service is not currently available to general gemini_python users, plans to provide this as a local calibration service are in place and expected for Future Enhancements.
As a matter of operations, reduce and the Recipe System depend upon the services of what is called the adcc, the Automated Data Communication Center. The adcc provides services to pipeline operations through two proxy servers, an XML-RPC server and an HTTP server. The XML_RPC server serves calibration requests made on it, and retrieves calibrations that satisfiy those requests from the Gemini FITS Store, a service that provides automated calibration lookup and retrieval.
The adcc can be run externally and will run continuously until it is shutdown. Any instances of reduce (and the Recipe System) will employ this external instance of the adcc to service a pipeline’s calibration requests. However, a user of reduce need not start an instance of the adcc nor, indeed, know anytihng about the adcc per se. If one is not available, an instance of the adcc will be started by reduce itself, and will serve that particular reduce process and then terminate.
This note is provided should users notice an adcc process and wonder what it is.
One enhancement long imagined is what has been generally termed ‘intelligence’. That is, an ability for either reduce or some utility to automatically do AstroDataType classification of a set of data, group them appropriately, and then pass these grouped data to the Recipe System.
As things stand now, it is up to the user to pass commonly typed data to reduce. As shown in the previous section, typewalk, typewalk can help a user perform this task and create a ‘ready-to-run’ @file that can be passed directly to reduce. Properly implemented ‘intelligence’ will not require the user to determine the AstroDataTypes of datasets.
The Fits Storage service will be delivered as part of a future release and will provide the calibration management and association features of Fits Storage: for use with the public release of the gemini_python data reduction package. This feature will provide automatic calibration selection for both pipeline (recipe) operations and in an interactive processing environment.