System-level management method of container in cloud calculating

Abstract

The invention provides a system-level management method of a container in cloud calculating. The method comprises the following steps that during a service i operation process, a service monitoring unit i sends a query request message of an effective container quantity to a container control unit according to a time interval, wherein the query request message corresponds to a service i; the container control unit returns a container quantity m to a service monitoring unit i; the service monitoring unit i compares whether the container quantity m is less than a target container quantity n, if the container quantity m is less than the target container quantity n, a difference value is set to be p, and an instruction of immediately creating p new containers is issued to the container control unit; and the container control unit distributes the created p new containers to the service i. The method has advantages that a container which is distributed to one service and is converted to be in an ineffective operation state can be timely discovered and the new available container is timely distributed to the service; based on that, when one service is started and during an operated whole life process, the containers with a constant quantity are always distributed and the service is guaranteed to be continuously operated; and a service capability, stability and reliability of a system are increased.

Description

technical field [0001] The invention belongs to the technical field of cloud computing, and in particular relates to a system-level management method for containers in cloud computing. Background technique [0002] In the existing cloud computing environment, the load balancing mode adopted is: add a proxy server between the front-end service and the background server cluster, whenever the front-end service needs resources, the front-end service sends a resource allocation request to the proxy server; the proxy server receives After receiving the request, according to the set load balancing strategy, a specific background server is selected from the background server cluster, and the selected background server provides resources to the front-end service. [0003] However, the inventor found in the process of realizing the present invention that the prior art has at least the following problems: [0004] During the entire life process of a front-end service being started and...

AI Technical Summary

Problems solved by technology

[0004] During the entire life process of a front-end service being started and running, the proxy server only assigns a fixed back-end server to the front-end service, and once the assigned back-end server fails, it is difficult for the proxy server to effectively and quickly know the back-end server If a fault occurs, a new background server will be reassigned for the front-end service, which will lead to interruption of the front-end service and reduce the service capability of the entire cloud computing system; and when the service capability of the entire cloud computing system is reduced, the cloud computing system will be The possibility of larger and more failures increases, which further affects system stability

Images