Advances in parallel computing have resulted in the growth of grid computing, a model that enables the integration of highly distributed and heterogeneous compute elements into a single, ubiquitous environment for solving computational problems. Often applied to large-scale efforts in fields such as structural biology, computational chemistry and high-energy physics, grid applications require infrastructure that supports the sharing and storage of increasing quantities of data. Grid initiatives are challenged to serve the growing demands of these data-intensive grid applications.;A data grid is a grid solution that integrates storage devices to provide a data service network for grid applications. Data grids house and serve data to the grid, and must be a robust, scalable, and seamless environment for both users and their computational jobs. The distribution of files across a network of storage devices can represent significant challenges to data management and delivery in a grid environment. A data grid is an essential component of a generic grid.;We will discuss three initiatives for the Advanced Computational Data Center Grid (ACDC Grid) that we have designed to address these issues and to provide a reliable data grid infrastructure. (Abstract shortened by UMI.)... |