Experiences Acquiring and Distributing a Large Scientific Database

Abstract

In recent years volumes of scientific data and of scientific databases have increased tremendously. Distributing a database across multiple sites is one possible solution to the problem of managing growing data. We address some practical aspects of distributing a large scientific database - the Sloan Digital Sky Survey (SDSS) database. As a preliminary, we also describe how we acquired the SDSS data and set up a mirror site. The architecture of our grid-based distributed database system is presented.