Design Pattern Based Distribution of Microservices in Cloud Computing Environment

Abstract

Cloud computing is a paradigm that has already evolved. Cloud computing moved widely to microservices from monoliths. The modular cloud application has gained attention for Microservices. Intensive network communication is required to call the interdependent microservices operating inside the cloud nodes. This research focuses on container-based microservices pre-distribution techniques and proposes two distribution strategies i.e. design pattern distribution and random distribution. The microservices are arbitrarily distributed to the available data centers in the random allocation method. While the microservices are clustered in the pattern distribution based on behavioral design patterns, which identify common contact patterns between entities. A custom-built modeling environment has been used to evaluate the proposed method. The findings revealed that the pre-distribution of microservices in accordance with the application architecture trend led to substantial less response time for the calls made to services hosted at geographically dispersed data centers.