OpenStack cluster nesting deployment method and system

Abstract

The invention provides an OpenStack cluster nesting deployment method and system, and the method comprises the steps: obtaining an arrangement template, carrying out the grammar correction processing of the arrangement template, obtaining a target Heat template, and determining a capacity expansion and contraction template; detecting a performance monitoring index of the internal cluster based on the capacity expansion and contraction template, and when the performance monitoring index meets a preset alarm condition, compiling a template type corresponding to the capacity expansion and contraction template into a template type corresponding to a capacity expansion and contraction execution template adaptive to the internal cluster so as to determine the capacity expansion and contraction execution template adaptive to the internal cluster; and transmitting the capacity expansion and contraction execution template into the internal cluster so as to execute capacity expansion and contraction operation on the internal cluster. According to the method and the device, the free deployment of the external cluster and the internal cluster is realized, or the external cluster and the internal cluster are simultaneously deployed, and the capacity expansion and shrinkage operation of the internal cluster by means of the external cluster is realized.

Description

technical field [0001] The present invention relates to the technical field of cloud computing, in particular to a method and system for nested deployment of OpenStack clusters. Background technique [0002] Existing OpenStack cluster deployment solutions are generally implemented in four ways, namely, automated cluster deployment based on Ansible, automated cluster deployment based on Kolla-Ansible, RDO (Remote Desktop Management Tool) automated cluster deployment, or manual deployment of clusters. [0003] However, the above methods cannot solve the problem of cluster nested deployment, in which, manually deploying each service component has the disadvantage of low deployment efficiency, and if the cluster scale is large, it will take a long time to deploy a complete set of Cluster; ROD is an official cluster deployment tool, but it can only deploy OpenStack clusters of the community version; Kolla-Ansible can deploy OpenStack clusters of different versions, but if you wan...

AI Technical Summary

Problems solved by technology

[0003] However, the above methods cannot solve the problem of cluster nested deployment. Manually deploying each service component has the disadvantage of low deployment efficiency, and if the cluster scale is large, it will take a long time to deploy a complete set of cluster; ROD is an official cluster deployment tool, but it can only deploy the community version of the OpenStack cluster; although Kolla-Ansible can deploy different versions of the OpenStack cluster, but if you want to implement cluster nested deployment, human intervention is also required. implementation, there are also defects of low deployment efficiency and cumbersome deployment

In addition, the aforementioned methods to achieve nested deployment of clusters have high technical requirements for developers

[0004] For the expansion and contraction of the cluster environment, it can be realized through the Ceilometer monitoring and metering system based on Heat. Ceilometer generally has poor response performance, and the message queue system (MessageQueue, MQ) is also used to push performance indicators. Therefore, If there are too many messages, it will be too late to process, so there is a shortcoming of not being efficient enough in the nested deployment cluster to perform expansion and contraction scenarios

The second method is to containerize the various components of OpenStack and use the Kubernetes application orchestration system to realize cluster environment monitoring and expansion and contraction. This method has the following disadvantages: On the one hand, additional deployment of Kubernetes services is required, which increases the Resource overhead; on the other hand, it is necessary to use Kubernetes services proficiently, and the technical threshold requirements for operation and maintenance personnel are relatively high

Images