Research And Development Of Real Time Incremental Backup Tool Platform For Large Structured Database

With the continuous development of the Internet,information technology is now closely bound up with our daily life.More and more enterprises are now accelerating the development of information technology and automation.Data storage,as the most important part of information technology,its security is becoming increasingly important.If there are not enough disaster recovery measures to protect a large-scale structured database,the severity of damage can be imagined once some accidents caused by data loss occur.At present,the backup tools of some large-scale structured database have defects such as low efficiency,weak timeliness and applicability.Therefore,we designed and developed a real-time incremental backup tool platform,which is based on file system changes monitor and does not rely on any large structured database.The platform is designed in layered structure.It is mainly divided into two functional layers,the file system monitor layer and the synchronization layer.The file system monitor layer is responsible for monitoring the real-time changes in physical files of the database.When it detects some changes of database files,it will notify the synchronization layer for synchronous operations.Synchronization layer will incrementally synchronize the changed file to the backup side,in order to achieve efficient real-time backup from the source side to the backup database.This paper will take DB2 as an example to analyze the structure of physical files in the database,then to build the overall framework of tools platform and implement it.We use Linux Inotify mechanism to implement the file system monitor layer and a mature synchronization tool,Rsync,is used to implement the synchronization layer.We found that the Rsync tool is not applicable for some cases in the tool platform,so we made some analysis and functional optimization to Rsync tool in order to enhance the efficiency and ensure the timeliness of tool platform.In the end of the paper,we made some systematic tests on the function and performance of the tools platform.The results indicate that our tool platform performs well and can effectively achieve real-time incremental backup to large structured database.