Huawei Technologies Co., Limited

Xiangping Wu

Long Li

Lin Hu

Peixing Peng

A cluster system self-organizing method including receiving, by a node A in a cluster system, a request message sent by an external node for requesting to join the cluster system, where the node A is one node in the cluster system, the node A and at least one other node in the cluster system mutually have a bidirectional neighbor relationship, and two nodes that have the bidirectional neighbor relationship mutually transmit information, separately determining, by the node A, whether a ratio of a space area that is of the node A and that is allocated in the cluster system to a space area that is of each bidirectional neighbor node of the node A and that is allocated in the cluster system meets a preset relationship, and establishing, by the node A, the bidirectional neighbor relationship with the external node when the preset relationship is met.

Cluster system self-organizing method, device, and cluster system

The present disclosure provides a method for managing a virtual machine cluster. The method includes: sending, by the first virtual machine, a first heartbeat message to at least two neighboring virtual machines in the N virtual machines, so that the at least two neighboring virtual machines detect the first heartbeat message, where a result of detecting the first heartbeat message is used to determine a status of the first virtual machine, and the first virtual machine establishes a neighbor relationship with the at least two neighboring virtual machines according to the information about the N virtual machines; and detecting, by the first virtual machine as a neighboring virtual machine of a second virtual machine in the at least two virtual machines, a second heartbeat message sent by the second virtual machine.

Method for processing virtual machine cluster and computer system

A bidirectional forwarding detection (BFD) method, apparatus, and system, where when a first virtual machine receives BFD packets from a second virtual machine, prediction duration and a predetermined quantity of sample time intervals are obtained. An eigenvalue is obtained according to the prediction duration and the predetermined quantity of sample time intervals, where the eigenvalue indicates a possibility of fault occurrence on an application in the second virtual machine. For an application that is running on the second virtual machine, the eigenvalue is compared with a preset fault determining criterion of the application, and whether the application is faulty is determined according to a comparison result.

Growing community of inventors

Shenzhen, China

Xiangping Wu