论文部分内容阅读
随着越来越多的机构采用集群计算技术来实现高性能计算(HPC)——地球、海洋和大气科学、地震数据分析等科学研究和药物研究、汽车设计模型、商业业务冒险分析等商业应用,集群计算技术已经进化为开发高性能计算系统的主要的方法。所有的这些应用的计算都是公认的复杂。能够有效地管理与这些应用密切相关的数据集驱动了现在的集群计算技术的发展。采用智能存储设备(OSD)大大简化了元数据服务器的工作量,并且使得系统的管理和效率都得到了很大的提高。集中描述一个新的存储体系结构──基于智能OSD的共享存储集群计算系统。
As more and more organizations adopt clustered computing to enable high performance computing (HPC) - commercial applications such as earth sciences, oceanography and atmospheric sciences, seismic data analysis and other scientific research and drug research, automotive design models, and commercial business risk analytics , Cluster computing technology has evolved as the primary method of developing high-performance computing systems. All of these applications are generally accepted as computationally complex. The ability to effectively manage data sets that are closely related to these applications has driven the development of today’s clustered computing technologies. The use of intelligent storage devices (OSDs) greatly simplifies the workload of the metadata server and enables significant improvements in system management and efficiency. Focus on describing a new storage architecture ─ ─ based on intelligent OSD shared storage cluster computing system.