Virtualization of Large-Scale Data Storage System to Achieve Dynamicity and Scalability in Grid Computing

Abstract

Data storage management is one of the most challenging issues for Grid resource management since large amount of data intensive applications frequently involve a high degree of data access locality. Grid applications typically deal with large amounts of data. In traditional approaches high-performance dedicated servers are used for data storage and data replication. This allows opportunistic grids to share not only the computational cycles, but also the storage space. This paper explains new mechanism for Dynamic and Scalable Storage Management (DSSM) in grid environments is proposed. The storage can be transparently accessed from any grid machine, allowing easy data sharing among grid users and applications. The concept of virtual ids that, allows the creation of virtual spaces has been introduced and used. The DSSM divides all Grid Oriented Storage devices (nodes) into multiple geographically distributed domains and to facilitate the locality and simplify the intra-domain storage management. Grid service based storage resources are adopted to stack simple modular service piece by piece as demand grows. To this end, we propose four axes that define: DSSM architecture and algorithms description, Storage resources and resource discovery into Grid service, Evaluate purpose prototype system, dynamically, scalability, and bandwidth, and Discuss results. Algorithms at bottom and upper level for standardization dynamic and scalable storage management, along with higher bandwidths have been designed.