A high-availability method and system for active-standby switching of control nodes based on galera

Abstract

The invention discloses a high-availability method and system for main / standby control node switching based on Galera, and relates to the field of cloud calculation. The system comprises a main control node, a standby node, a main monitoring and sending module, a standby monitoring and sending module, a main storage module, a standby storage module, a main timing and executing module and a standby timing and executing module. The main monitoring and sending module monitors and sends heartbeat information of the main control node and receives heartbeat response information sent by the standby monitoring and sending module. The standby monitoring and sending module monitors the heartbeat information of the standby control node, receives the heartbeat information and sends the heartbeat response information to the main monitoring and sending module. The main storage module and the standby storage module receive and store cluster information. The main timing and executing module calculates whether the heartbeat response information receiving process of the main monitoring and sending module is overtime. The standby timing and executing module calculates whether the heartbeat information receiving process of the standby monitoring and sending module is overtime. By adopting the high-availability method and system, the integrity of data and the normal work of clusters are ensured when switching is carried out between the main control node and the standby control node.

Description

technical field [0001] The present invention relates to the field of cloud computing, in particular to a Galera-based high-availability method and system for active-standby switching of control nodes. Background technique [0002] In the cloud computing data center, the control node is the core of the entire cluster in cloud computing by managing the computing nodes, storing the cluster status, configuring data, and dynamically scheduling cluster resources through the control node. [0003] When the control node is down, the cluster corresponding to the control node will be in the state of no central node management, making some functions of the cluster unable to use normally; when the data on the control node is damaged, the cluster corresponding to the node Status and configuration data will be lost, and even user business data will be lost. Therefore, the control node affects the normal operation of the cluster. [0004] At present, in order to avoid the failure of the c...

AI Technical Summary

Problems solved by technology

[0009] When using MySQL for asynchronous replication, the data is first written to the master control node, and then copied to the backup control node. When the master control node is down, some data has been written to the master control node, but it has not Standby control node. At this time, the standby control node is upgraded to the master control node. The data that is not copied to the standby control node is lost, which affects the normal operation of the cluster.

Images