This research presents a network analytics structure and procedure designed to support real-time processing of large-scale data. The structure consists of multiple data adaptive nodes and data analytic clusters arranged in a distributed network. The data analytic nodes within the clusters employ a peer-to-peer networking mode and a load balancing mechanism, enabling the clusters to dynamically expand and contract as needed. The structure utilizes a streamlined, flow-based analytic processing approach among the data analytic nodes, facilitated by an incident mechanism. By leveraging this network analytics structure, real-time analysis and processing of enormous network data, including fault monitoring, statistics, troubleshooting, and diagnosis, can be achieved. The structure allows for fine-grained analysis of network data and supports dynamic expansion of structure functionalities, allowing users to define their own analytic requirements. Additionally, the structure's distributed architecture reduces reliance on individual hardware performance, enabling efficient processing of complex logic for network data analysis. The structure and procedure support various types of processing logics, thereby reducing the expertise required from developers.

Network analytics, Real-time data processing, Distributed structure, Scalability, Peer-to-peer networking, Load balancing, Fine-grained analysis