Testing Spatial Data Deliverance in SQL and NoSQL Database Using NodeJS Fullstack Web App

During the last decade, we saw an explosion of geospatial data being produced. Most of which coming from GPS-enabled devices available for general consumers. The large amount of geotagged data coined the term `Geospatial Big Data', indicating the semi-structured and unstructured nature of such data. SQL relational databases have been known in the past to handle geospatial data very well. However, the abundance of geospatial big data pushed forward the need for NoSQL database which is expected to perform better in terms of handling and storing geospatial big data. This paper discusses the quantitative comparison of performance between the SQL (i.e., PostGIS) and NoSQL (i.e., MongoDB) databases in handling geospatial big data. A NodeJS-based angular-framework web app was developed to test the real-world performance of MongoDB and PostGIS in handling a large amount of simulated geospatial data. A different number of points were generated for testing the geospatial data storing and loading capability of both the databases. The test was conducted by comparing the result of XHR (XML HTTP Request) of both databases in each case. The result showed that NoSQL database, i.e. MongoDB, performs better in loading big geospatial data compared to traditional SQL database using PostGIS.