Performance Evaluation of MongoDB and PostgreSQL for Spatio-temporal Data

Several modern day problems need to deal with large amounts of spatio-temporal data. As such, in order to meet the application requirements, more and more systems are adapting to the specificities of those data. The most prominent case is perhaps the data storage systems, that have developed a large number of functionalities to efficiently support spatio-temporal data operations. This work is motivated by the question of which of those data storage systems is better suited to address the needs of industrial applications. In particular, the work conducted, set to identify the most efficient data store system in terms of response times, comparing two of the most representative of the two categories (NoSQL and relational), i.e. MongoDB and PostgreSQL. The evaluation is based upon real, business scenarios and their subsequent queries as well as their underlying infrastructures, and concludes in confirming the superiority of PostgreSQL. Specifically, PostgreSQL is four times faster in terms of response time in most cases and presents an average speedup around 2 in first query, 4 in second query and 4,2 in third query in a five node cluster. Also, we observe that the average response time is significantly reduced at half with the use of indexes almost in all cases, while the reduction is significantly lower in PostgreSQL.