作者: Riccardo Zappi , Elisabetta Ronchieri , Alberto Forti , Antonia Ghiselli
DOI: 10.1007/978-1-4419-8014-4_19
关键词:
摘要: In production data Grids, high performance disk storage solutions using parallel file systems are becoming increasingly important to provide reliability and speed I/O operations needed by High Energy Physics analysis farms. Today, Storage Area Network commonly deployed at Large Hadron Collider centres, such as GPFS Lustre reliable, native POSIX in fashion. this paper, we describe StoRM, a Grid middleware component, implementing the standard Resource Manager v2.2 interface. Its architecture fully exploits potentiality offered underlying cluster system. Indeed, it enables encourages use of protocol (i.e. ”file://”) allowing managed Element improve job efficiency accessing. The running on worker node can perform direct access StoRM if were local disk, instead transferring from Elements disk.