A novel polyglot data mapper for an E-Commerce business model

In the current scenario E-Commerce sites are replete with content. However, they are facing major challenges in areas of scalability and availability. Entity-attribute data designs are very complex and provide less flexibility to the data layer of the application. Relational databases are better for transactions and ACID guarantees but they prove to be a limiting factor in providing linear scalability for large volumes of data. Schema less No SQL stores are better when scalability is a requirement. Every No SQL data store has its own pros and cons; hence simply replacing legacy RDBMS systems with a single No SQL does not suffice. Rather, one needs to have a combination of different data stores for different modules of a business model. However, addition of multiple data stores in the application brings new challenges and increases complexity of maintaining the application. In this paper, we present a hybrid data store architecture using a polyglot data mapper for E-Commerce aiming to achieve better scalability whilst ensuring simplistic maintenance of the application infrastructure.