Task multi-level decompose in grid-VGE

In order to integrate kinds of data, models and applications in virtual geography environment (VGE), we propose the grid-VGE based on the technology of grid. With regard to the specific applications of grid-VGE, the tasks submitted by users can be divided into several sorts, such as science computing, I/O species, data management, simulating and modeling. With the development of VGE, more and more applications will join the grid-VGE, but they may not belong to any species that has been listed up. Of course, many tasks may be covered by several species. When a user submits a complex and onerous task, the task needs to be decompounded to assign to different locations for implementation. Based on the theoretic of Bernstein by layer upon layer, we decompose the task into many sub-tasks, which will be distributed to certain nodes based on the upper capability hierarchy. And these sub-tasks will cooperate to complete the whole task. In an addition, a status-word will be added to every sub-task which is used to record the status of the sub-task implementation. In the last, we will give an example. The user can ramble and interact with each other in the three-dimension visional scene and can use kinds of peripheral equipments to hence the sense of immersion and interact each other in real time. The experiment results show that decomposition and distribution of tasks regarding grid-VGE are correct and effective for the cooperative implementation of the tasks.

[1]  Ian T. Foster,et al.  The Anatomy of the Grid: Enabling Scalable Virtual Organizations , 2001, Int. J. High Perform. Comput. Appl..

[2]  Lingyun Yang,et al.  Conservative Scheduling: Using Predicted Variance to Improve Scheduling Decisions in Dynamic Environments , 2003, ACM/IEEE SC 2003 Conference (SC'03).

[3]  Ian T. Foster,et al.  Conservative Scheduling: Using Predicted Variance to Improve Scheduling Decisions in Dynamic Environments , 2003, ACM/IEEE SC 2003 Conference (SC'03).

[4]  Shoubao Yang,et al.  Decentralized Computational Market Model for Grid Resource Management , 2003, GCC.

[5]  Ian T. Foster,et al.  Application experiences with the Globus toolkit , 1998, Proceedings. The Seventh International Symposium on High Performance Distributed Computing (Cat. No.98TB100244).

[6]  Kavitha Ranganathan,et al.  Decoupling computation and data scheduling in distributed data-intensive applications , 2002, Proceedings 11th IEEE International Symposium on High Performance Distributed Computing.

[7]  Ian T. Foster,et al.  Homeostatic and tendency-based CPU load predictions , 2003, Proceedings International Parallel and Distributed Processing Symposium.