Grid Environment Based On Multiple Data Sources Data Transmission

Data-intensive scientific and engineering applications often require the efficient transfer of terabytes or even petabytes of data in wide-area, distributed computing environments. Therefore, how to transfer large amounts of data efficiently and securely in the grid environment has became the key issue of grid computing.MFTP, a new data transferring tool based on multiple replicas in the grid environment is one important part of MASSIVE (Multidisciplinary Applications-Oriented Simulation and Visualization Environment)grid platform. MASSIVE aims to use Grid technology to establish an enabling environment for distributed simulation and visualization of large-scale scientific and engineering research. MASSIVE grid platform is built on the top of Globus Toolkit and aims to offer users a grid-enabled CFD (computational fluid dynamics) and CSM (computational solid mechanics) problem solving.The MFTP proposed in this paper adopts a new approach to improve performance further by using replica-level parallelism transfer in Grids. It's instruction showes as follows. MFTP is invoked with the logical identifier of a grid data file, then the MFTP will query the Replica Location Services to find as many corresponding replicas as possible. Subsequently, MFTP starts to contact each replica source to start one or more sub-transferring using security authentication, parallel transfer, partial transfer of GridFTP. In this section, workload placement algorithms decide when and which replica should transfer which part (offset and length) of the source file. Data is gathered from these source replicas and eventually combined together to a single file. As while as improving the performance of data transfer, MFTP applys the GSI and Kerberos security authentication of the GridFTP, so it also has the security of the GridFTP. Corresponding experiments show that MFTP improves the data transfer rate and reliability on Grids.