ITR: Computing Grid for Analysis of Data from PHENIX at RHIC

该项目处于信息技术和核物理的交界处。我们将在石溪和布鲁克海文国家实验室开发和应用一个双站点数据网格来分析相对论重离子对撞机PHENIX实验中的重离子碰撞。RHIC的碰撞产生数以千计的粒子，导致需要分析的大量数据。为了研究这种罕见的大角度胶子散射过程，以及核介质如何影响这一过程，我们必须对数亿个这样的复杂事件进行筛选。我们将使用Globus Collaboration开发的软件工具，允许在两个计算机设施之间进行无缝数据分析，这两个计算机设施包括许多计算机和TB的磁盘存储。作业将自动定向到所请求的数据驻留在磁盘上的设施，并将结果返回给用户。我们将利用其他人在信息技术方面的许多发展，但我们带来了一个独特的组合，包括真实数据、渴望的用户和一个尖端的数据网格应用程序，其中包含两个硬件非常相似的邻近站点的简单演示网格。在“战斗”条件下测试和优化这个原型网格将提供适用于未来大规模全球数据网格开发的信息，以及来自PHENIX的物理结果。

This project is at the interface between Information Technology and Nuclear Physics. We will develop and apply a two-site data grid at Stony Brook and Brookhaven National Laboratory to analysis of heavy ion collisions in the PHENIX experiment at the Relativistic Heavy Ion Collider. Collisions at RHIC produce many thousands of particles, resulting in enormous volumes of data to analyze. In order to study the rare process of gluon scattering at large angle, and how the nuclear medium affects this process, we must sift through hundreds of millions of these complex events. We will use the software tools developed by the Globus collaboration to allow seamless data analysis across two computer facilities comprising many computers and terabytes of disk storage. Jobs will be automatically directed to the facility where requested data are disk resident, and results returned to the user. We will take advantage of many developments by others in information technology, but we bring to the table a unique mix of real data, eager users, and a cutting edge data grid application with a simple demonstration grid of two nearby sites with very similar hardware.Testing and optimizing this prototype grid under "battle" conditions will provide information applicable to development of large-scale global data grids for the future, as well as physics results from PHENIX.