Data Consistency in Large Scale Applications

Abstract

One of the challenges in large scale application is data consistency. In this paper, we have studied a few of the algorithms which are used to tackle this issue. We have also conducted some interview sessions with software experts and analyze their judgment regarding this issue. We have tried to analyse this topic based on two major software architecture. Mircoservice and monolithic architecture. Microservice is a trending software architecture for large scale application nowadays. It has many advantages along with some drawbacks. There are many services that communicate with each other to do a single operation. So, maintaining consistency is a big concern here. There are some ways to ensure consistency. Generally, microservice-based application use eventual consistency model. But this model has a chance of data loss. In some cases, data loss can be threatening to the business. To avoid this situation strong consistency model can be used. But algorithms that supports strong consistency comes with performance degradation of software compare to eventual consistency model.