ANALYSIS

Time series data consists of information collected over time, often at regular intervals, which can quickly accumulate in large volumes. To analyze and present this data effectively, it needs to be stored in a way that is easy to access. Database Management Systems (DBMSs) are commonly used for this purpose. There are various types of DBMSs, each with distinct advantages and disadvantages that involves different trade-offs regarding their features. In this study, we compare the performance of two different DBMSs for managing time series data. The first is PostgreSQL, a widely used relational DBMS enhanced by the Timescale DB extension designed for time series. The second is Mongo DB, a leading NoSQL system that offers built-in support for time series data. Investigation was made to ascertain which of these DBMS is better suited for time series data by analyzing their query execution times. Two databases were setup using sample time series dataspecifically, publicly available weather. A series of test queries were conducted simulate real world scenarios and measure their execution times The results are analyzed on a query-by-query basis to highlight performance between two systems, with PostgreSQL outperforming MongoDB on some queries (by over two orders of magnitude) while MongoDB was faster on others (by more than 30 times in one case). It was concluded that certain queries and their related real-world applications may favor one DBMS over the other, based on how well the query structure aligns with the system’s strength. Additional factor was discussed that may affect each DBMSs query execution efficiency and consider potential improvements.