专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明申请

US20100223618A1 SCHEDULING JOBS IN A CLUSTER 失效
标题翻译：在集会中安排工作
公开(公告)号：US20100223618A1
公开(公告)日：2010-09-02
申请号：US12713740
申请日：2010-02-26
申请人： ZHI WEN FU , Zi Ming Hu , Peng Liu , Jie Yang
发明人： ZHI WEN FU , Zi Ming Hu , Peng Liu , Jie Yang
IPC分类号： G06F9/46
CPC分类号： G06F9/50 , G06F9/5044 , G06F9/505 , G06F9/5077 , H04L67/1002
摘要： There is provided a method and system for scheduling a job in a cluster, the cluster comprises multiple computing nodes, and the method comprises: defining rules for constructing virtual sub-clusters of the multiple computing nodes; constructing the multiple nodes in the cluster into multiple virtual sub-clusters based on the rules, wherein one computing node can only be included in one virtual sub-cluster; dispatching a received job to a selected virtual sub-cluster; and scheduling at least one computing node for the dispatched job in the selected virtual sub-cluster. Further, the job is dispatched to the selected virtual sub-cluster based on characteristics of the job and/or characteristics of virtual sub-clusters. The present invention can increase the throughput of scheduling effectively.
摘要翻译：提供了一种用于调度集群中的作业的方法和系统，所述集群包括多个计算节点，并且所述方法包括：定义用于构建所述多个计算节点的虚拟子集群的规则; 基于规则将簇中的多个节点构建成多个虚拟子集群，其中一个计算节点只能包含在一个虚拟子集群中; 将所接收的作业调度到所选择的虚拟子集群; 以及为所选择的虚拟子集群中的所分配的作业调度至少一个计算节点。此外，基于作业的特性和/或虚拟子集的特征将作业调度到所选择的虚拟子集群。本发明可以有效地提高调度的吞吐量。

2. 发明授权

US08631410B2 Scheduling jobs in a cluster having multiple computing nodes by constructing multiple sub-cluster based on entry and exit rules 失效
标题翻译：通过基于进入和退出规则构建多个子集群，在具有多个计算节点的集群中调度作业
公开(公告)号：US08631410B2
公开(公告)日：2014-01-14
申请号：US12713740
申请日：2010-02-26
申请人： Zhi Wen Fu , Zi Ming Hu , Peng Liu , Jie Yang
发明人： Zhi Wen Fu , Zi Ming Hu , Peng Liu , Jie Yang
IPC分类号： G06F9/46 , G06F11/00
CPC分类号： G06F9/50 , G06F9/5044 , G06F9/505 , G06F9/5077 , H04L67/1002
摘要： There is provided a method and system for scheduling a job in a cluster, the cluster comprises multiple computing nodes, and the method comprises: defining rules for constructing virtual sub-clusters of the multiple computing nodes; constructing the multiple nodes in the cluster into multiple virtual sub-clusters based on the rules, wherein one computing node can only be included in one virtual sub-cluster; dispatching a received job to a selected virtual sub-cluster; and scheduling at least one computing node for the dispatched job in the selected virtual sub-cluster. Further, the job is dispatched to the selected virtual sub-cluster based on characteristics of the job and/or characteristics of virtual sub-clusters. The present invention can increase the throughput of scheduling effectively.
摘要翻译：提供了一种用于调度集群中的作业的方法和系统，所述集群包括多个计算节点，并且所述方法包括：定义用于构建所述多个计算节点的虚拟子集群的规则; 基于规则将簇中的多个节点构建成多个虚拟子集群，其中一个计算节点只能包含在一个虚拟子集群中; 将所接收的作业调度到所选择的虚拟子集群; 以及为所选择的虚拟子集群中的所分配的作业调度至少一个计算节点。此外，基于作业的特性和/或虚拟子集的特征将作业调度到所选择的虚拟子集群。本发明可以有效地提高调度的吞吐量。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式