Research On Dynamic Deployment Of Micro Service In Cloud Environment

With the rapid development of Internet and cloud computing,enterprise applications migrate from a single computer to the cloud computing platform gradually.The high availability and resiliency of the cloud computing environment drives the Web application architecture model from the original standalone architecture to the popular micro service architecture.This change effectively reduces the development time of enterprise applications,and improving the fault tolerance,scalability and on-demand capacity of applications.However,with the complexity of the business and the increasing number of micro-services,there are many problems with the dynamic deployment of micro services in the cloud environment:1.Due to the loosely coupled nature of the micro service architecture,each micro service can be developed using the team's development language,independently deployed and expanded,but it is unavoidably dependent on the hardware and software environment at the time of deployment.2.In order to reduce the coupling among micro-services,generally use lightweight messaging mechanism(such as RESTful style HTTP API way to communicate).This kind of remote communication mechanism may have some problems,such as the response is not timely,the service is busy or the target is unavailable,because the micro service architecture has a smaller granularity and a business request usually invokes multiple micro services.3.Micro service is an independent module of small particle size in physics,but the logical dependency exists,and usually rely on multiple levels.In order to protect the user's experience,service providers will generally deploy a certain amount of redundant services when the load is high,and reduce the amount of services to save resources when the load is small.But the multi-level dependent micro service under this mechanism is difficult to control,when to increase or decrease the number of micro services and how many micro services show be increased or decreased is difficult to determine,which may cause the system to frequently increase or decrease the dynamic deployment of micro services.However,the frequent dynamic deployment will reduce the stability of the whole system,and then SLA can not be guaranteed,and it will lead to a waste of resources at the same time.This paper analyzes the characteristics of the cloud platform,the key technologies and difficulties in the deployment of micro services deployment are studied,the algorithm proposed in this paper can combine the advantages of cloud computing and containerization technology to dynamically deploy the micro service in a reasonable,effective and automatic way without user awareness,in the event of micro-service failure,the algorithm can detect and repair it in a timely manner,and ensures the stability of the system and reduces the cost of enterprises as much as possible.The main work of this paper is as follows:1.In order to reduce the dependence of the micro service on the software environment,using the Docker container to package micro services and its runtime environment.On this basis,the micro service pool model is proposed.2.The pool model guarantees a certain amount of redundancy in the micro service instance,the fault will be promptly discovered and quickly repaired when part of the micro service is busy or unavailable.This solves the problem that the local micro service is unavailable.3.In order to solve the problem of frequent dynamic deployment of micro service in cloud computing environment under different load conditions,a dynamic deployment method of micro service oriented to SLA is proposed.A dynamic description model of micro-service is built,endow the micro service and its dependencies with dependency weight,and offer a dynamic update mechanism of the dependency weight,it can accurately locate the time required for dynamic deployment by monitoring changes in the dependent weights,and then,according to the dependency weights,the corresponding number of layers are deployed.This method improves the service guarantee of SLA and reduces the risk of default.Finally,the validity of the proposed model and algorithm is verified by experiments,and analysis and comparison of the actual deployment results.In theory,the proposed model and algorithm can be used as a technical support for dynamic deployment of micro services in cloud environment,and it can reduce the difficulty of business operations and the cost of practical use.