Architecture perspective of NoSQL : user experience and scalability of cassandra and MongoDB

Abstract

The enormous amount of data stored today by web applications and non-web applications alike are becoming alarming as a result of slow development of storage technologies required to cope with the scale and agility challenges. This is very typical of highly ordered data that needed to be accessed by multiple applications concurrently. As developers and database administrators made effort to cope with users data, new technologies emerged to solve the limitations of traditional database designs and provide more features to enhance scalability and user exxperience. This thesis examined the traditional relational database management systems and NoSQL database technology. The aim of this work was to test the two technologies by comparing a relational database with two NoSQL databases and analyze how NoSQL technology is providing solutions to the scalabilty and performance limitations of relational databases. As a result of this thesis, a relational database and two NoSQL databases were implemented to hold the data of a Twitter clone application. The results of the comparison revealed that MongoDB can better handle complex queries than MySQL because of its simple and flexible schema. MySQL on the other hand would better serve in simple search queries because of its logical data structure. Cassandra exhibit the most data availability due to data partitioning over sets of nodes.