Research On Context-based Microservice Adaptive Invoking Technology

With the continuous updating and iteration of the software system,the software system becomes more and more huge and bloated,which brings great inconvenience to the later development and maintenance of the system.In order to solve this problem,the industry has proposed the concept of microservice architecture.The microservice architecture splits the system into multiple microservices,each of which implements independent business functions.However,since the microservices are deployed in a cluster manner,the same microservice corresponds to multiple microservice instances,which brings difficulties to the response of the microservice request and the invoking of the microservice.In this paper,the microservice invoking technology is deeply studied from the aspects of service request queue adjustment and microservice instance selection.The purpose is to improve the accuracy of microservice instance selection and microservice execution efficiency,and further realize high-quality microservice application..Based on the analysis and research of microservices,context and message queue theory,this paper makes an in-depth study on context-based microservice adaptive invoking technology.Firstly,a microservice adaptive invoking framework is designed.The framework gives the whole process of adaptive invoking of the microservice instance,which achieves the accurate invoking of the microservice instance,the correct transmission of the user request and the real-time adjustment according to the actual load of the server.A message queue-based microservice request queue adjustment algorithm is proposed and applied to the microservice adaptive invoking framework to achieve the asynchronous communication between microservices and the decoupling between microservices.The load condition of the server adjusts the order of requests inside the microservice request queue,which improves the efficiency of service execution.At the same time,a context-based microservice instance selection algorithm is proposed,which takes into account the QoS information and user context information in the Web service,and also considers the server load information in the microservice cluster,combined with QoS,user context and the server loads these three kinds of context information to dynamically obtain the user preference weight vector,and considers the influence of each context on the user's required service instance from the historical history of the user access,so that the selection result balances the subjective and objective factors and improves the accuracy of the service instance selection and service execution efficiency.Next,the context-based microservice adaptive invoking framework is implemented and the microservice request queue adjustment algorithm and microservice instance selection algorithm proposed in this paper are compared with other related algorithms.Finally,the effectiveness and usability of the context-based microservice adaptive invoking framework and algorithm studied in this paper are verified by a practical application case of microservice deployment and invocation.