The Research And Implementation Of Backup And Recovery For Large Scale Archived Streams

With the development of information technology, in many domains there is a new type of high-speed, massive, dynamic data, which is called Data Stream, The application of Data Stream is very broad, one of common applications is to archive and to store the data for users to research, analyse, inquire, this kind of data which needs storing is called Archived Stream. The archived stream with high commercial and research value is very precious, if it is lost or not available, the consequences would be unthinkable, So how to ensure these precious archived streams available? How to store and preserve these large scale archived streams? This requirements require us to study the backup and recovery's technology of the large scale archived streams.In the background of the massive archived streams' database management system, this paper makes a in-depth study in the database technology,backup and recovery technology of the massive archived streams, and massive data loading technology, and designs a backup and recovery system which is adapted well to complex hardware and network environment of massive archived stream's database management system, and influences the online system smally with high backup and recovery performance. Specifically, in this paper the following aspects of work are carried out:According to requirements of archived streams' backup and recovery, we design the architecture of archived stream's backup and recovery system, which is a complete backup and recovery system architecture including data backup, recovery, storage, and historical data query.We do the physical and logical design for the massive archived streams' database management system, which make its database structure to adapt to requirements of large scale archived streams' backup, store and management system.Discarding the traditional database backup technology, We use the method of backuping source data,This method backuping simply and quickly, backuping data smaller and interfering the online system smaller is feasible, efficient and lifts the large scale archived streams' backup performance greatly.We develop massive archived streams' loading technology, and apply it in the recovery of large scale archived streams, This technology combines volume data loading, district exchange mechanism, the multi-level parallel scheduling, and other technologies, which reaches GB-per-second high-speed parallel loading, increases the speed of recovery greatly and meets the needs of large scale archived streams.