Performance Analysis of Geospatial Serverless Computing for Geospatial Big Data Analysis

Nowadays, real-world applications are primarily concerned with collecting the physical characteristics that correspond to various geographical phenomena, and information plays a critical role in the analysis and forecast of various occurrences that occur. Real-time analysis is used in a variety of application domains, including traffic flow monitoring, healthcare monitoring, and so on. This research study has considered the existing geospatial serverless computing framework and proposed an analytical model to enable the geospatial serverless platform to compute and operate at different workloads and the tradeoff between the cost and performance are then determined by the users’ preferences. Here, the accuracy of the model is validated in Amazon Web Service (AWS) Lambda environment and the model calculations are shown based on the performance metrics by including average response time, probability of cold start as well as the average amount of functional instances in the steady state by using the queuing theory. Here, the performance modelling is carried out in the geospatial serverless environment for different workloads, which results in delivering better performance at a minimal cost.