公开(公告)号：US20160036924A1

公开(公告)日：2016-02-04

申请号：US14551078

申请日：2014-11-24

Applicant: Microsoft Technology Licensing, LLC.

Inventor： Lokesh Srinivas Koppolu ,  David A. Dion ,  Abhinit Kumar ,  Amitabh P. Tamhane ,  Andrea D'Amato ,  Daniel Upton ,  Elden T. Christensen ,  Nicholas Maliwacki

IPC: H04L29/08 ,  H04L12/26

CPC classification number: H04L67/16 ,  G06F9/5088 ,  G06F11/1484 ,  G06F11/2028 ,  G06F11/2035 ,  G06F2201/81 ,  H04L41/0654 ,  H04L41/0896 ,  H04L41/5054 ,  H04L43/04 ,  H04L67/1025 ,  H04L67/322

Abstract: Embodiments provide workload processing for clustered systems. In an illustrative, non-limiting embodiment, a computer-implemented method may include identifying a server as an active node of a cluster; assigning a workload to the server in response to the identification; determining, after the assignment, that the server is no longer an active node of the cluster; calculating, in response to the determination, a probability that the server is capable of continuing to execute the workload; and deciding, based upon the probability, whether to allow the workload to remain assigned to the server.

Abstract translation: 实施例提供了集群系统的工作负载处理。 在说明性的非限制性实施例中，计算机实现的方法可以包括将服务器标识为集群的活动节点; 响应于所述标识向所述服务器分配工作负载; 在分配之后确定服务器不再是集群的活动节点; 响应于所述确定，计算所述服务器能够继续执行所述工作负载的概率; 并且基于概率来决定是否允许工作负荷保持分配给服务器。

公开(公告)号：US10609159B2

公开(公告)日：2020-03-31

申请号：US14551078

申请日：2014-11-24

Applicant: Microsoft Technology Licensing, LLC

Inventor： Lokesh Srinivas Koppolu ,  David A. Dion ,  Abhinit Kumar ,  Amitabh P. Tamhane ,  Andrea D'Amato ,  Daniel Upton ,  Elden T. Christensen ,  Nicholas Maliwacki

IPC: H04L29/08 ,  H04L12/26 ,  G06F11/30 ,  G06F9/50

Abstract: Embodiments provide workload processing for clustered systems. In an illustrative, non-limiting embodiment, a computer-implemented method may include identifying a server as an active node of a cluster; assigning a workload to the server in response to the identification; determining, after the assignment, that the server is no longer an active node of the cluster; calculating, in response to the determination, a probability that the server is capable of continuing to execute the workload; and deciding, based upon the probability, whether to allow the workload to remain assigned to the server.