Big Data Access Control for Cloud-Native Hadoop Environments

The rise of Big Data necessitates robust access control for platforms like Hadoop. While traditionally deployed on physical servers within trusted networks, Hadoop is increasingly migrating to cloud-native, containerized environments. This transition introduces significant security challenges, as the compromise of a single container can potentially expose other resources. Existing Big Data access control models, designed for traditional configurations, often lack the necessary flexibility for dynamic cloud-native environments. This research proposes a usage control-based model to secure privileged access to Big Data and its processing within containerized environments. The paper analyzes existing access control solutions and explores Hadoop architectures in cloud-native deployments. It then presents a model leveraging usage control and multi-step authorization to address these evolving security needs. The proposed approach enhances traditional access control by incorporating organizational context and approval workflows for sensitive operations. It mitigates the risks associated with unbounded privileges and rogue container deployment by enabling real-time, reactive policy enforcement. Unlike existing models, this solution offers dynamic adaptability, fine-grained control, and improved resilience against insider threats, making it particularly well-suited for securing Big Data in modern, distributed environments.