An Optimized Kappa Architecture for IoT Data Management in Smart Farming

Agriculture 4.0 is a domain of IoT in full growth which produces large amounts of data from machines, robots, and sensors networks. This data must be processed very quickly, especially for the systems that need to make real-time decisions. The Kappa architecture provides a way to process Agriculture 4.0 data at high speed in the cloud, and thus meets processing requirements. This paper presents an optimized version of the Kappa architecture allowing fast and efficient data management in Agriculture. The goal of this optimized version of the classical Kappa architecture is to improve memory management and processing speed. the Kappa architecture parameters are fine tuned in order to process data from a concrete use case. The results of this work have shown the impact of parameters tweaking on the speed of treatment. We have also proven that the combination of Apache Samza with Apache Druid offers the better performances