A Grid Scheduling Algorithm Based on Resources Monitoring and Load Adjusting

Grid can integrate massive idle resources into a high-performance supercomputer, which is good choice for resolving the complicated engineering optimization problems. However, the heterogeneous, distributed and dynamic characters of the grid resources makes tasks scheduling are very difficult in the engineering optimization. A grid scheduling algorithm which is based on resources monitoring and load adjusting is presented for tasks scheduling in the grid environments. This method uses monitoring information of the resources to select the powerful resources and finish the initial distribution. Load adjusting will prevent the imbalance adaptively based on the feedback information form the selected resources. Simulations have been carried out and more detail analysis has been done for the algorithm. At last, a sample of injection plastic optimization was conducted on grid by using the scheduling model, and results prove that the proposed method is reasonable and effective.