Download now Free registration required
Grid computing is often viewed as a vehicle to realize the promise of distributed computing in large-scale heterogeneous environments. Simply put, it enables the virtualization of key resources like CPU, memory, disk and storage spread across disparate systems as a single managed entity. Grid computing solutions are often tasked with the basic functional requirements like Managing and virtualizing a distributed set of resources, scheduling tasks (in parallel) across a set of compute nodes and maintaining a desired level of quality of service. A detailed analysis of grid applications reveals that such applications can be divided into two categories based on their function - long-running process grids and high-performance real-time grids. The first category involves grids used in areas such as overnight Value-at-Risk (VaR) calculations in investment bank, or life sciences research, or the SETI project, or in the area of Computational Fluid Dynamics (CFD). Most recent popular trend is the synergy between Service Oriented Architectures (SOA) and grids. As grid computing continues to grow as a popular paradigm in most IT organizations, there is an obvious need to expand most pilot grid implementations to a larger scale to support several enterprise business operations. But, such expansion plans are often stymied by data latency bottlenecks that cripple grid processes. This paper explores the detailed analysis of the grid computing for high end performance.
- Format: PDF
- Size: 347.6 KB