Research on High Availability Technology of Data Based on Autonomous Platforms
In order to solve the problem of traditional high availability clusters relying on shared storage,this article im-plements a dual write technology for file system data on the primary and backup machines in a dual node situation by adding an I/O filtering model to the file system driver.When the application on the host writes file system data,the I/O da-ta can be synchronized in real-time to the backup machine.When the host goes down,the application on the backup ma-chine remounts the file system on the backup machine,original file system data is not lost.