An Efficient E-Commerce Design by Implementing a Novel Data Mapper for Polyglot Persistence

Abstract

E-commerce applications are under tremendous pressure of performance with the increase of data, users and cutthroat competition. They need to optimize their designs to handle scalability and availability and at the same time ensure consistency. Traditional data layers were built with single type of data store and focused on very few database qualities. Traditional RDBMS was very effective if only consistency was desired. NoSQL stores are very effective if either availability or scalability was desired. But in an e-commerce application, we need all the properties all together. Hence, there is a need of hybrid data layer or polyglot persistence in the data layer. We need a mix of SQL and NoSQL both in the data layer of e-commerce. Implementing polyglot persistence needs to take care of appropriate insertion and retrieval. Data should be inserted in their respective database, and retrieval should also be done accurately. Polyglot persistence requires a data mapper to do this job. Here, we have designed a novel data mapper with adapts to the situation and executes the commands.