Overview

The flow of Astro-Ddata is illustrated by Figs. 1-3. It begins with the receipt of the original data from ISAS. The data are ingested into the Astro-Ddatabase management system where they are logged and complete observations are identified. The data are converted to FITS data files--the SF & HKF--followed by the concatenation of complete observations. The SF is then cleaned and filtered to produce the FSF FITS file for scientific analysis. The scientific analysis itself is the next stage, producing derived data products. All data received and all data product files created by the pipeline will be stored in the Astro-Darchive. When data loose their proprietary status, they will be moved from the Astro-Ddatabase to the HEASARC archive for free access.

The processing system will be highly modular and available for general distribution. The system will be capable of being run with some modules or files missing, thereby protecting it from complete failure should, e.g., the attitude file be unavailable. The software will be built in such a way as to facilitate repairs as much as possible. For example, an Astro-Dinvestigator will be able to run a simple procedure in his or her home institution to repair a broken or out-of-date module. Calibration updates will also be implemented in this way. Configuration control will be an integral feature of the system.

Three versions of the processing system will be produced, all built from the same software components. These versions are:

• The Quick Analysis system will be designed first and is intended to provide the instrument groups with the means to analyze data as promptly as possible. It will also be used by duty scientists to check data.
• The Standard system will be what GOs use to process and analyze their data. It will be differ from the quick analysis system in that more corrections and more data-selection criteria will be applicable. GOs will be able to choose which analysis package, such as IRAF/PROS, IDL or XANADU they wish to use.
• The Pipeline system will be the same as the standard system but will run automatically in batch mode. The data products and the results of the scientific analysis will be channeled automatically into the Astro-Darchive.

In cooperation with ISAS, software and hardware in the US and Japan will be as close to identical as possible. To this end, the software is being written in ANSII-standard C and FORTRAN to run on machines with the Ultrix operating system. We will ensure that the software will also run under SunOS and under as many other flavors of Unix as possible. Software will be written as much as possible in a system-independent fashion.

Each analysis task will be steered by a command-driven database operating system. This will use a DBMS to select observation intervals, to access tables of data and to write back to tables of results. The DBMS will record the state of the processing and allow archiving to a mass storage device.