Serverless Architecture for Big Data Analytics

Abstract

Big data is a phrase that describes the large quantity of data, it would be structure, semi structure and unstructured. In the present industry data is indispensable for the business organization. The Big Data initiatives and technologies are used to analyze this massive amount of data for gaining insights which may help in making strategic decisions. For example, data size is increasing day by day like petabyte, Exabyte, zettabyte, yottabyte and more. That is why it is going to tough and complex to manage this large scale of data. In practical, there are many challenges to process and compute this data like server management, storage, clustering, algorithm deployment, etc. As everything is done by manually, so it is hard to design the exact architecture for data analysis in the Cloud. Serverless computing is a mechanism to provide pay-per-use backend services to clients. A serverless provider lets users create and deploy code without worrying about operating and managing servers. In this paper, we present serverless architecture for big data analytics, also we show how to implement, maintain, and governance of a serverless big data application on AWS (Amazon Web Service). In addition to it, we will demonstrate the difference between traditional data analytics and big data analytics in a serverless system.