Distributed Continue Data Protection Solution

With the development of information technology in production and life, it strongly stimulates the growth of massive data. So information security and real-time have attached great importance by various sectors. The continuous protection of data is gradually raised to the agenda.Continuous data protection is a major breakthrough in the field of data protection. The traditional data protection programs just focus on the cyclical backup of the data, only manage to create, control and restore a single time point of data backup. But the continuous data protection can captures all changes of data to get the recovery at any previous time point. Although there are some data synchronization and backup products in current market, it lacks protection backup device for personal data. The project is set in the continuous data protection project of Lenovo Company. It will provide a distributed protection solution for personal data.The paper does some analysis on the realization model of the existing data protection program and then it puts forward the basic calculation model based on metadata-driven. Rabin-based fingerprint algorithms to document differences is adopt to complete the quick backup and rapid recovery from the three stages of data protection life including data generation, data analysis the data construction. In the meantime, the paper describes the staged event-driven task scheduling module in order to resolve highly-concurrent server problems. In the aspect of network framework, the system uses a dual-bus network design to address the long-server connections and overloading problems. The paper also designs the module of version controlling, so as to complete the history version management and the version inter-sharing and co-editing.This article introduces the meaning and content of the project at first, and then makes detailed analysis on the module design and realization. Finally, it does a summary and certain outlook to the project.