From Relational Databases to NoSQL Databases: Performance Evaluation

Abstract

In nowadays applications, the amount of data in the database grows exponentially. So, the DBMS must process these huge amounts of data as fast as possible. The main aim of this study is to prove that NoSQL databases process big data faster than relational database. The changing in applications, user and infrastructure characteristics, mostly of the Web 2.0 domain and cloud platform, led to explosion of data sources and massive workloads. These huge amounts of data have raised the problems of storage and usability of data as the usual Relational Database Management Systems (RDBMS) were unable to handle the exponentially growing data on a single in house server (vertical scalability). Furthermore, the required time to process these big data is an issue. The study implements prototype which verifies performance argument. The performance evaluation compares the insertion and retrieval speeds between MongoDB as NoSQL database and MySQL as relational database. The benchmarking performed shows that MongoDB is faster than MySQL in the most of scenarios we chose, particularly when we deal with huge amount of data.