Research And Implementation Of Distribution And Transferring In A Distributed Storage System Based On Grid

To share distributed computer resources all over the world, and easier to user the leisure CPU time and storage space, the idea of grid and grid computing come out to handle these problems. We wish that we could access the resources such as, computing resources, storage resources, science instruments, etc, as using electric power. There need a kind of middle software to join all computer nodes to cooperated together to finish some kind of project, because one computer will do computing work, but it may handling data on another computer.There are different kinds of storage systems in the grid environment. These storage systems have been created in response to specific needs for storing and accessing large datasets. They each focus on a distinct set of requirements and provide distinct services to their clients. Some storage systems pay more attention to high performance access to data, so they should use parallel data transferring and striping to improve performance. Some other storage systems focus on supporting high volume usage and using data set replication and local caching to divide and balance server load performance.The new proposed data transfer protocol need to define how to transfer data rapidly; it should be compilable with other common protocols and fit for transferring data on the Internet, it should include partial data transferring because clients usually need only part of data of a file; third transferring for user can control data transferring between two data servers; restarting data transfer, and continue transferring after the connection broke; it also should maintenance the security of data transferring.The research work described in this paper is research and implementation of distribution and transferring in a distributed storage system based on grid. We analyzed the mechanism, module and implantation of data transferring over the grid, include data module, mechanism of data distribution. We implement a storage grid in our lab, and developed a web service compliant with WSRF on grid environment to access data server. Finally we do several performances testing about cohesion of grid base on WSRF and SOA in order to improve our system.