Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.stecf.org/conferences/adass/adassVII/hulberts.html
Дата изменения: Mon Jun 12 18:51:52 2006
Дата индексирования: Tue Oct 2 00:09:38 2012
Кодировка:
Поисковые слова: п п п п п п п п п п п п п п п п р п р п р п р п р п р п р п р п р п р п р р п п р п п р п п р п п р п п р п п р п р п п р п р п п р п р п п р п р п п р п р п п р п
|
Next: REMOT: A Design for Multiple Site Remote Observing
Up: Dataflow and Scheduling
Previous: Data-flow for the ESO Imaging Survey (EIS)
Table of Contents -- Index -- PS reprint -- PDF reprint
Astronomical Data Analysis Software and Systems VII
ASP Conference Series, Vol. 145, 1998
Editors: R. Albrecht, R. N. Hook and H. A. Bushouse
S. Hulbert
Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218
Abstract:
The routine processing of STIS data in the calibration ``pipeline" requires
an interrelated set of software tasks and databases developed and maintained
by many groups at STScI. We present the systems-level design of this calibration
pipeline from the perspective of the STIS instrument group at STScI. We
describe the multiple interfaces to the pipeline processes that the STIS group works
across in an effort to provide for the routine calibration of STIS data. We
provide a description of the pipeline processing while treating the actual processing
software as a collection of black boxes. We describe the systems engineering
requirements levied by and against the STIS group needed to carry out day-to-day
calibration of STIS observations.
The functional view of the STIS calibration pipeline (Figure 1)
illustrates the interactions between the external world and the pipeline itself.
In this case, there are two sets of external users: observers (including GO and GTO)
and the STIS group at STScI. The two groups of users are similar in that they both
receive data products consisting of raw and calibrated STIS data from the
calibration pipeline system. Additionally, the STIS instrument scientists are
responsible for calibrating STIS and feeding these calibration parameters back to
the pipeline system. (Note: the arrows that are used in all three figures indicate the flow of information between groups.)
To demonstrate the number of details that must be managed to run a
calibration pipeline for STIS we identify distinct entities within the Space
Telescope Science Institute with which the STIS group interacts in the course of
keeping the STIS calibration pipeline up and running. Borrowing from the
object-oriented paradigm, we designate these entities as ``objects".
The Object Interaction Diagram (OID) of the STIS calibration pipeline
(Figure 2) shows in detail the ``objects" within STScI that the STIS group
interacts
with in the process of ensuring that STIS data are properly calibrated. These
objects consist of three distinct classes: Operations, Database and Tools.
The Operations class consists of subsystems (usually controlled by distinct
management groups) that actually handle STIS data on a daily basis. The Database
class consists of databases containing myriad details of STIS operations which are
cared for by the database designers and operators. The third class of object in
this model, Tools, consists of groups of programmers with the responsibility of
producing specialized software tools. The OID concepts used in this description of the calibration pipeline have been adapted from Morris, et al. (1996).
Figure 1:
Functional View of STIS Calibration Pipeline.
|
Figure 2:
Top Level Object Interaction Diagram of STIS Calibration Pipeline
|
Missing from the OID of the STIS calibration pipeline is the explicit interaction between the STIS group and each object in the model of the STIS calibration pipeline. The OID shown in Figure 3 gives, as an example, the detailed description of the interface between the STIS group and the Tools class object, the
STSDAS programming group. The STSDAS group is responsible for crafting the highly specialized calibration code used to process STIS data. This code runs not only in the pipeline but also as stand-alone tasks. Special attention should be paid to the interactions (pair of arrows) labeled ``Troubleshooting". A critical part of negotiating the implementation of
STISspecific features of the pipeline is the ability and willingness of the STIS
group and the relevant operations group, programming group or database group to
identify and resolve problems in a timely manner.
Figure 3:
Object Interaction Diagram of the STIS Group Interface with the STSDAS Group
|
To assist in the management of the STIS portion of the calibration pipeline, the STIS group is subject to constraints imposed by the systems engineering
of the pipeline process. For example, the STIS group:
- maintains ICD-19 which is a description of the data formats for all
STIS data including specifications for the header keywords. This
information is stored in the Keyword Database (KWDB).
- maintains ICD-47 which is a description of the STIS calibration
reference file formats. This information is stored in the Calibration
Database System (CDBS).
- performs cross-subsystem testing of software changes affecting STIS.
- participates in the biweekly Pipeline Working Group consisting of
representatives from all of the calibration pipeline ``objects" (groups).
- provides documented specifications via instrument science reports
(ISRs), technical instrument reports (TIRs) and the problem
reporting (OPR) system.
References:
Morris D., Evans, G., Green, P. & Theaker, C. 1996. Object Oriented Computer Systems Engineering.
(New York: Springer), 70
© Copyright 1998 Astronomical Society of the Pacific, 390 Ashton Avenue, San Francisco, California 94112, USA
Next: REMOT: A Design for Multiple Site Remote Observing
Up: Dataflow and Scheduling
Previous: Data-flow for the ESO Imaging Survey (EIS)
Table of Contents -- Index -- PS reprint -- PDF reprint
payne@stsci.edu