期刊文献+

基于Portlet的高性能计算应用集成组件 被引量:3

High Performance Computing Application Integration Toolkits Based on Portlet
原文传递
导出
摘要 高性能计算Web应用社区经常需要封装不同类型的应用,封装过程需开发或调整源代码。因此,社区管理员不能快速满足应用封装需求,特别是作业参数相对复杂的应用。通过分析不同作业管理系统的作业提交和管理命令、不同学科领域的高性能计算应用程序的使用方式、通用作业描述语言的结构和语义,本文提出了一种基于portlet的高性能计算应用集成组件,设计和实现了应用需求描述模板、作业提交页面动态渲染引擎、作业参数存储和填充方法,从而提供了基于WEB页面的应用封装、动态生成作业提交Web页面和作业描述并提交作业的功能,并且支持以历史作业的输入数据填充作业提交页面的功能。分析及实例表明,该组件能够有效提高应用封装的响应速度,用户能够方便的提交作业和重用复杂的作业参数,具有良好的可移植性和扩展性。 In high-performance computing Web communities, we need to encapsulate many kinds of application softwares into Web portals or gateways and need to develop or adjust source codes to add theses softwares.Therefore, it is difficult for administrators of Web communities to meet requests for adding softwares quickly and easily, especially softwares that require lots of complicated parameters. By analyzing requirements and formats of job submission and management commands provided by different job management systems for clusters and supercomputers, using methods of lots softwares used in different disciplines and research areas, syntax and structure of generic job submit description language, we proposed high-performance computing applications integration toolkits(HPC-AIT) based on portlet. In HPC-AIT, we designed and implemented a template for describing requirements of application softwares, a dynamics rendering engine for job submiting Web pages, and methods for saving and filling complicated parameters. The HPC-AIT provided application packaging service based on the Web page, dynamic generation of job submitting web pages, automatic generation of a job description, and submitting jobs to computing resources. It also supported to fill the job submitting page by the data from the specific history job. The analysis and examples show that, administrators can package application quickly and easily, users can submit jobs and reuse the parameters from the specific history job. In addition, it is good at portability and scalability.
出处 《科研信息化技术与应用》 2017年第2期18-30,共13页 E-science Technology & Application
基金 国家重点研发计划项目课题"国家高性能计算环境构建与资源提升关键技术"(2016YFB0201404) 中国科学院院级科研项目"中国科技云建设工程"课题"超算资源池建设"(XXH13501-03)
关键词 高性能计算 应用封装 作业提交和管理 参数填充 计算网关 high-performance computing application encapsulation job submitting and management fillingparameters computing portal
  • 相关文献

参考文献3

二级参考文献23

  • 1Thomas S.High Performance Computing[EB/OL].(2010-03-01).http://en.wikipedia.org/wiki/High-performance_computing.
  • 2Foster I,Zhao Yong.Cloud Computing and Grid Computing 360-Degree Compared[C] //Pros.of IEEE GCE'08.Austin,USA:[s.n.] ,2008.
  • 3Anjomshoaa A,Drescher M.Job Submission Description Language Specification Version 1.0[EB/OL].(2005-11-07).http://www.gridforum.org/documents/GFD.56.pdf.
  • 4黄凯 徐志伟.可扩展并行计算技术、结构与编程[M].北京:机械工业出版社,2000..
  • 5樊建平 陈明宇.高性能计算机研究的现状与展望[J].信息技术快报,2003,(5).
  • 6TOP500 Supercomputing Sites.http://www.top500.org.
  • 7PVM:Parallel Virtual Machine.http://www.epm.ornl.gov/pvm/pvm_home.html.
  • 8The Message Passing Interface (MPI) standard.http://www-unix.mcs.anl.gov/mpi/.
  • 9OpenMP:Simple,Portable,Scalable SMP Programming.http://www.openmp.org.
  • 10Jin H.ChinaGrid: Making Grid Computing a Reality. Digital Libraries: International Collaboration and Cross-Fertilization . 2005

共引文献30

同被引文献19

引证文献3

二级引证文献20

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部