Performance Analysis of Self-adaptive Policies in Containerized Microservices

Nowadays applications tend to be designed following a microservice based architecture, deployed on containers and handled by orchestration platforms on the Cloud. This increasing attention to containerized microservices is due to the high modularity, flexibility and scalability of their architecture. Container technologies running on orchestration platforms are skyrocketing in popularity among developers and in deploying all kinds of microservices and applications such as smart vehicles, IoTs and fog/edge computing. Therefore, efficient management of Cloud resources at run time through self-adaptive clusters has become of prime importance in Cloud computing. This paper presents a Stochastic Petri Net modelling approach to analyze the performances of self-adaptive policies in containerized microservices. We illustrate the effectiveness of our approach through a set of experimental results.