Cluster processing method and device, electronic equipment and computer readable storage medium

Abstract

The invention provides a cluster processing method and device, electronic equipment and a computer readable storage medium, and relates to the field of cloud computing. The method is applied to a cloud server, and comprises the following steps: when a conversion instruction for converting a preset native cluster into an edge cluster is received, generating a first component and a second componentbased on a preset edge cluster management component; modifying the parameters of the native cluster interface based on the first component to obtain a modified native cluster interface, and sending the second component to each node in the native cluster, so that the second component deploys an agent component in each node, and takes a cluster composed of the deployed nodes as an edge cluster; andthus, each node in the edge cluster performs data interaction with the modified native cluster interface in the cloud server through the proxy component. According to the invention, automatic conversion from the native cluster to the edge cluster is realized.

Description

technical field [0001] The present application relates to the technical field of cloud computing, and in particular, the present application relates to a cluster processing method, device, electronic equipment, and computer-readable storage medium. Background technique [0002] Kubernetes is an application for managing containerization on multiple hosts in a cloud platform. It supports automated deployment, large-scale scalability, and application containerization management. A native cluster deployed on Kubernetes can be called a Kubernetes native cluster, and an edge cluster deployed on Kubernetes can be called a Kubernetes edge cluster. [0003] In the existing technology, there are two solutions for converting a native Kubernetes cluster into a Kubernetes edge cluster: [0004] Solution 1: Deploy a tailored version of the edge cluster by relying on the Kubernetes deployment method [0005] An obvious feature of the Kubernetes edge cluster of this type of solution is to...

AI Technical Summary

Problems solved by technology

[0012] 1) Because the edge device is not like a cloud server, it may be just a smart box without a public network IP, and resources are generally limited, especially CPU (central processing unit, central processing unit) and memory, without tailoring Kubernetes The major core components and edge devices cannot fully meet the operating environment of the major core components of Kubernetes;

[0013] 2) Without tailoring the core components of Kubernetes, you can only expand your own edge technology solutions on Kubernetes

[0017] 1) After tailoring the functions of the core components of Kubernetes, adding its own edge technology solutions, and following the deployment scheme of Kubernetes, although it is relatively easy to deploy, it has a fatal flaw: it is not fully compatible with Kubernetes, and with the iteration of Kubernetes And development, the problem will become more and more serious, not to mention the introduction of new features of Kubernetes;

[0020] Converting a Kubernetes cluster to an edge Kubernetes cluster by hand or script is cumbersome and cannot be implemented in batches in a production environment

The script can achieve the goal indirectly, but it lacks the corresponding flexibility, linkage with other components of the Kubernetes cluster, and is strongly dependent on the basic environment. It is extremely error-prone as the environment changes, and the compatibility and scalability are poor.

[0021] To sum up, Solution 1 will cut out the major core components of Kubernetes, add its own edge technology solutions, and follow the Kubernetes deployment solution to achieve the purpose of deploying Kubernetes edge clusters. However, with the enrichment of solutions and the development of Kubernetes, the Kubernetes Compatibility and resource occupation of edge devices will become worse and worse as time goes by; Solution 2 can achieve the goal in disguise through manual and scripting, but lacks flexibility, scalability and Kubernetes linkage , and highly error-prone

Images