Abstract:Using the P2P model to scheduling the grid job among the multi-cluster
it not only enhances the grid system scalability and fault-tolerant
but also turns each cluster into an autonomous scheduling unit and completes their job cooperatively. The system integrates the Globus grid middleware system and the OpenPBS
and increases assisted scheduling services
which makes grid job migrate among the multi-cluster and eases effectively the load of job in a single cluster.