Performance Evaluation of NoSQL Multi-Model Data Stores in Polyglot Persistence Applications

Abstract

NoSQL data store systems have recently been introduced as alternatives to traditional relational database management systems. These data stores systems implement simpler and scalable data models that increase the performance and efficiency of a new kind of emerging complex database application. Applications that model their data using two or more simple NoSQL models are known as applications with polyglot persistence. Usually, their implementations are complex because they must manage and store their data using several data store systems simultaneously. Recently, a new family of multi-model data stores was introduced, integrating simple NoSQL data models into a single unique system. This paper presents a performance evaluation of multi-model data stores used by an application with polyglot persistence. In this research, multi-- model datasets were synthesized in order to simulate that application. We evaluate the performance of benchmarks based on a set of basic database operations on single model and multimodel data store systems. Experimental results show that in some scenarios multi-model data stores have similar or better performance than simple model data stores.