Closed loop management system framework of acquisition operations and maintenance based on data homology management

Abstract

In order to solve the problems existing in the management of information collection, operation and maintenance, such as slow fault detection, difficult location, lack of predictability, difficult initiative, emphasis on technical quality, light of indicators, lack of process, lagging behind, difficulty to precipitate knowledge, low efficiency and so on, a data acquisition and analysis platform based on data homology management based on Hadoop is built. The platform provides a basic supporting framework for big data analysis, supports online and offline distributed data processing capabilities, and provides a unified data access and storage interface to form a closed-loop management system framework. The data loading subsystem is also implemented, which provides different effective data acquisition schemes for the collection of multi-source heterogeneous operation and maintenance logs. The correlation analysis subsystem of operation and maintenance logs is designed and implemented, providing the ability of feature extraction, clustering, on-line classification and correlation analysis for off-line and on-line data analysis of logs. The framework is constructed and designed based on OSGI framework. Finally, the key technical indicators of the system are tested. It is proved that the system can effectively promote data homology, quality and business operation and maintenance efficiency.