论文部分内容阅读
ARC(Advanced Resource Connector)CE(Computing Element)是Nordugrid开发的网格中间件中的对计算资源管理的组件。ARC CE相对于其他的对等计算资源网格组件,具有轻量级,易管理,认证简单,易扩展的特性,因此越来越广泛地被应用于各类网格应用平台上。高能物理实验ATLAS具有数据量大,计算量大的特点,采用了网格计算的方式整合了位于全球上百个计算中心的15万个CPU核进行协同计算。随着ATLAS Run2阶段开始,其所需计算资源急速增长,因此ATLAS在积极探索使用云计算资源,超算资源(HPC),志愿计算资源等。如何将这些异构的计算资源整合到ATLAS传统的网格计算平台,成为一个急需解决的重要问题。ARC CE由于所具有的易扩展和轻认证的属性,于是成为这一解决方案的核心组件。本文将以志愿计算为例,阐述如何利用ARC CE将动态的、不可信的志愿计算资源平台整合到现有的ATLAS网格计算环境,实现对用户和网格服务的双透明性。
ARC (Advanced Resource Connector) CE (Computing Element) is a component of computing resource management in grid middleware developed by Nordugrid. Compared with other peer computing resource grid components, ARC CE is more and more widely applied to various grid application platforms because of its lightweight, easy to manage, simple authentication and easy expansion. High-energy physics experiment ATLAS has the characteristics of large amount of data and large amount of computation, and uses grid computing to integrate 150,000 CPU cores located in hundreds of computing centers in the world for collaborative calculation. ATLAS is actively exploring the use of cloud computing resources, hyper-computing resources (HPC), volunteer computing resources, etc., as the ATLAS Run 2 phase begins and its computing resources are rapidly growing. How to integrate these heterogeneous computing resources into ATLAS traditional grid computing platform has become an urgent problem that needs to be solved urgently. ARC CE has become the core component of this solution because of its scalability and light certification attributes. In this paper, volunteer computing is taken as an example to illustrate how ARC ARC can be used to integrate dynamic and unreliable volunteer computing resource platforms into the existing ATLAS grid computing environment to achieve dual transparency for users and grid services.