Data Handling And Maintaining Data Consistency In Scalable Replicated Micro-services

Abstract

Monolithic Architecture helps to develop server-side enterprise applications. But, it views as a “big ball of mud”. That indicates monolithic architecture has many drawbacks. Introduction of cloud based microservice architecture can solve these kind of drawbacks. Microservice architecture helps to scale an application. Most of the applications write less data than reading of that data. Scaling of read model and write model separately is very important. But, scaling applications using microservice architecture is very hard. Further, applications cannot simply use a local ACID (Atomicity, Consistency, Isolation, Durability) transaction. Read part is scaling to more replicas. Thus, maintenance of the data of all replicas in same level is important. Replication of read model and maintenance of data consistency which would provide experimental insight still needs to be developed. To bridge this gap, development of an architecture based on messaging with RabbitMq publish/ subscribe method, event sourcing and Command Query Responsibility Segregation (CQRS) with axon framework is used in this study. Evaluation of this architecture was done by replication of the read model using Docker and Docker-compose. Further, we have analysed data consistency in our experiments.