Cloud platform management node control method, device and cloud platform server cluster

Abstract

This application discloses a cloud platform management node control method, device, equipment, computer-readable storage medium, and cloud platform server cluster, which breaks through the problem of poor reliability of the traditional single management node or dual management nodes, through multiple management nodes The server builds a high-availability cluster, which improves the reliability of the system, and compared with the management node deployed in the form of a virtual machine, there is no problem of performance attenuation; at the same time, each management node server is mapped to a unified storage device to store the management node data. Compared with the traditional two-node disk mirroring mode, it can not only realize the storage of a larger amount of data, but also avoid the data split brain problem of saving data in the disk mirroring mode. Therefore, the cloud platform management node control method provided by this application has the advantages of high reliability and high performance, and breaks through the problems and bottlenecks of the management node deployment method in the prior art.

Description

technical field [0001] The present application relates to the field of cloud platform technology, and in particular to a control method, device, equipment, computer-readable storage medium, and cloud platform server cluster of a cloud platform management node. Background technique [0002] The cloud computing platform has become more and more mature in recent years. In the development of the national economy, the construction of information systems in various industries has evolved from the traditional physical machine model to the cloud platform model to support customers' key businesses. The cloud platform generally deploys a cloud operating system on a physical server, and hosts the Guest OS (system in a virtual machine) and customer applications and data on the cloud operating system. The types of nodes inside the cloud platform are divided into management nodes and computing nodes. The computing node runs the guest OS that hosts the customer. It is the node that hosts t...

AI Technical Summary

Problems solved by technology

[0006] If two management nodes are deployed, such as schemes 2, 4, and 7, although the management node has a dual-machine hot backup function, when the two management nodes of the cloud platform are all down, the system still has great risks, and During the dual-machine backup process, the backup data is transmitted through the network, and the data between the active and standby management nodes may be out of sync due to network abnormalities

[0007] If the management node is deployed in the form of a virtual machine, such as solutions 5, 6, and 7, compared with the management node deployed in the form of a physical machine, the CPU / memory / disk performance will be attenuated. If the overall scale of the cloud platform is large, There will be a situation where the management node is stuck

[0008] The sixth deployment scheme deploys the management node in the form of a virtual machine placed on the shared storage. If the management node runs abnormally on a certain computing node, other computing nodes can use HA to pull up the virtual machine, and the overall system The reliability is relatively high, but the management node is deployed in the form of a virtual machine, there will be performance degradation problems

Images