The rapid proliferation of Internet of Things (IoT) devices has revolutionized several industries by facilitating automated functionalities. Additionally, it empowers us to experience seamless connectivity and interaction with our surroundings. However, this widespread deployment of less secure IoT devices has also brought multifold security challenges, such as Distributed Denial of Service (DDoS) attacks. This attack may completely crash the victim’s online services and systems or make them unavailable to legitimate users. Several detection methods have been proposed in the literature. However, these methods are unable to provide a real-time detection solution for dynamic and high-speed IoT traffic-based attacks. In this paper, we propose a distributed classification approach for detecting DDoS attacks induced by IoT traffic in real-time, named by, DCA-IoT. This study centers around two key aspects: first, identification of the most suitable dataset for implementing the DCA-IoT approach, involving extensive dataset characterization and comparison. Second, development of an efficient DCA-IoT approach using distributed Machine Learning (ML) techniques, empowered by the high-performance Apache Storm stream-processing framework. We deployed the most efficient model on the Apache Storm cluster for analyzing incoming traffic streams quickly. The experimental results reveal that the proposed GBM-based distributed classification model achieves an accuracy of 99.73%.