ADASS XII Conference

Enabling technologies

---

P8.1 Efficient Distribution of Computational Load on a Bewoulf-Like Cluster

Luca Fini, Marcell Carbillet (Osservatorio Astrofisico di Arcetri)

The CAOS Application Builder is a Graphical Programming Environment which allows the building of complex simulation applications for Adaptive Optics systems, by putting together elementary blocks [1,2,3]. The resulting simulation programs are often very heavy in computational needs and could be profitably run on bewoulf-like clusters, provided the computational load can be efficiently distributed on the CPU's. In the paper we describe a project to provide the CAOS Application Builder with software tools which allow the user to optimize the distribution of blocks on a multi cpu machine and show a few preliminary results.

[1] L.Fini, et. al. in ADASS-X ASP Conference Series, Vol. 238, 2001. F. R. Harnden Jr., F. A. Primini, and H. E. Payne, eds. pp. 253-256
[2] M.Carbillet, et. al. in ADASS-X ASP Conference Series, Vol. 238, 2001. F. R. Harnden Jr., F. A. Primini, and H. E. Payne, eds. pp. 349-352.
[3] S.Correia, et. al. in ADASS-X ASP Conference Series, Vol. 238, 2001. F. R. Harnden Jr., F. A. Primini, and H. E. Payne, eds. pp. 404-407

P8.2 Source code management and software distribution using Open Source technologies

Martin Bly (Rutherford Appleton Lab, UK) Alasdair Allan (University of Exeter, UK) Tim Jenness (Joint Astronomy Center, HI)

The Starlink Software Collection (SSC) runs on three separate platforms and contains approximately 130 separate software items, totalling over 6 million lines of code. Distribution of such large software systems and installation at multiple remote sites has always been problematic due to the complex web of inter-dependences such systems invariably generate.

The rise of the Open Source movement has brought standard tools into common use to cope with such large and complex tasks. The Redhat Package Manager (RPM) software is one such which is available for many platforms. We have shown it is possible to automate the distribution and installation of the Starlink Software using RPM. We anticipate that this will vastly simplify installation and package management for System Administrators who must support the SSC in production data processing environments.

P8.3 SIRTF Mosaicker

David Makovoz, Khan Iffat

We present a software system for image coaddition/mosaicking that is being developed for the SIRTF mission. SIRTF mosaicker features the use of the drizzle interpolation technique, robust outlier detection based on spatial and temporal filtering, and fast direct plane-to-plane coordinate transformation. It is designed to interface with other tools developed at SSC, such as pointing refinement and overlap consistency, which will improve the quality of the mosaic images.

P8.4 MacOSX for Astronomy

F. Pierfederici, N. Pirzkal, R. Hook

MacOSX is the new version of the Unix based Macintosh operating system. It features a sleek, high performance, display PDF user interface, sitting on top of a standard BSD UNIX subsystem. Consequently, this OS empowers users with a broad range of applications previously non available on a single system such as Microsoft Office and Adobe Photoshop, as well as legacy X11-based scientific tools and packages (IRAF, SuperMongo, MIDAS, and Skycat). This combination of a modern GUI layered on top of a familiar UNIX environment paves the way for new, more flexible and powerful astronomical tools being developed while assuring compatibility with already existing, older programs. In this paper, we outline the strengths of the MacOSX platform in a scientific environment, Astronomy in particular, and point to the numerous astronomical software packages available for this platform; most notably the SciSoft collection.

P8.5 The Fasti Project

C. Baffa, V. Biliotti, A. Checcucci (INAF - Osservatorio di Arcetri) V. Gavrioussev (IRA - CNR) S. Gennari, E. Giani, F. Lisi (INAF - Osservatorio di Arcetri) G. Marcucci (Firenze University) M. Sozzi (IRA - CNR)

Fasti is a controller architecture originally developed for fast infrared astronomical array detectors, and intended to be powerful and extendible. It is suitable to be used with both DRO and CCD detector and it is also well suited for very fast optical detectors, as those used in Adaptive Optics. In the framework of LBT project, a L³CCD version is in development.

P8.6 The USNO-B Catalog

David Monet, Stephen Levine (USNO Flagstaff)

The USNO-B catalog presents positions, proper motions, magnitudes in various optical passbands, and star/galaxy estimators for 1,036,366,767 objects derived from 3,633,655,848 separate observations. The data were obtained from scans of 7,435 Schmidt plates taken for the various sky surveys during the last 50 years. A brief discussion of various will be presented, but the actual data are available from www.nofs.navy.mil and other sites after September 2002.

P8.7 Infrared-Array-Camera Images from the Space Infrared Telescope Facility

Russ Laher, Jason Surace, Heidi Brandenburg, Mehrdad Moshir

The Infrared Array Camera (IRAC), one of the science instruments on NASA's soon-to-be-launched Space Infrared Telescope Facility (SIRTF), has four simultaneous-imaging, focal-plane-array detectors with optical filters covering different near-infrared spectral regions: 3.6, 4.5, 5.8 and 8.0 m. IRAC digital images of the celestial sky will be computer-processed in several stages of a production pipeline for instrument-artifact removal and scaling to absolute intensity units. Ancillary uncertainty and pixel-condition-flag images will also be generated for each processed IRAC image. Special processing of calibration data will be done prior to applying it in the production processing. Telescope-pointing data will be separately processed and used to assign a sky position and orientation to each image. Time-sequences of images will be processed to create pixel maps of unwanted latent-image artifacts. Images overlapping the same sky region will be co-added to mitigate noise, put together to form sky mosaics much larger than IRAC's footprint on the sky, and further processed to yield intensity point-source information on the celestial objects that are imaged. Pixel maps of outliers in the image data will also be generated. Reduced data from IRAC's four infrared will be merged to facilitate scientific analysis. The data processing will be done at the SIRTF Science Center, first in real time and then in subsequent processing episodes to further refine the data products. Selected versions of the data products will be archived and made accessible to astronomers worldwide