Facilities for Remote Execution of High Performance Applications in Astronomy

The field of Astronomy deals with a large amount of observational and simulated data. Algorithms and software can process this data, revealing information about star dynamics, the center of galaxies, or even the structure of the Universe. Virtual observatory (VO) is a strategy to take advantage of distributed computer systems, sharing resources for data storage, processing and analysis. This work is a VO effort for remote execution of high performance applications in Astronomy. The VO was developed using the Python programming language with the Django web framework and the Celery distributed task queue. The platform offers parallel versions of the Friends-of-Friends algorithm for astronomical objects classification, and algorithms for image restoration using wavelet filtering. Users can set their execution parameters and files or use the ones provided by the platform. At the end of each execution, they can download the output file and a log file.