SQL vs NoSQL: A Spatial Benchmark for Selected NoSQL and Relational Database Management Systems

GIS applications are used to capture, store, manipulate, analyse, manage and present geographical data (Longley, 2005). These applications require functional and spatially enabled databases to operate. Relational database management systems (RDBMS) have been at the forefront in this field for an excess of 40 years. Over this time period, the demands on database management systems (DBMS) have changed, especially regarding the recent proliferation of Big Data and spatial data. The relational model is not perfect and often struggles to manage large, unstructured and highly related data. NoSQL DBMS have increased in popularity over the past two decades as a potential answer to some of the short comings of RDBMS. This investigation presents a scalable spatial database performance benchmark comparing two RDBMS (Oracle and PostgreSQL) with a document DBMS (MongoDB) and a graph DBMS (Neo4j). Findings show that the RDBMS are overall superior although MongoDB scales better than RDBMS when data is retrieved from disk output