首页> 外国专利> Method and system for job scheduling in a distributed data processing system using identification of optimal network topology

Method and system for job scheduling in a distributed data processing system using identification of optimal network topology

机译:使用最佳网络拓扑的标识的分布式数据处理系统中的作业调度方法和系统

摘要

To provide an automatic and optimized selection of network topology for distributed scheduling jobs on a computer having a network topology that has been modified. A selection automatically and optimization of the network topology, starts from a desired number of the current topology and additional connections. In this manner, the method, for example, when compared to a simple ring network, provide higher convergence speed for modified consensus algorithm. The method makes use of the so-called Small World network. Small World network, when compared to other network architectures, it is more robust to perturbation. Recommended embodiment, to match the workload increases heterogeneous distributed computing environment, highly expandable, it provides the workload scheduling system. In order to distribute network traffic and jobs among multiple computers, the modified average consensus algorithm was used. .FIELD 6
机译:为具有已修改的网络拓扑的计算机上的分布式调度作业提供网络拓扑的自动和优化选择。网络拓扑的自动选择和优化从所需数量的当前拓扑和其他连接开始。以这种方式,例如,当与简单的环形网络相比时,该方法为改进的共识算法提供更高的收敛速度。该方法利用了所谓的小世界网络。与其他网络体系结构相比,Small World网络具有更强的抗干扰能力。推荐实施例,为配合增加工作量的异构分布式计算环境,高度可扩展,它提供了工作量调度系统。为了在多台计算机之间分配网络流量和作业,使用了改进的平均共识算法。 .FIELD 6

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号