The AppLeS parameter sweep template: user-level middleware for the grid
Proceedings of the 2000 ACM/IEEE conference on Supercomputing
A taxonomy and survey of grid resource management systems for distributed computing
Software—Practice & Experience
Self-adaptive applications on the grid
Proceedings of the 12th ACM SIGPLAN symposium on Principles and practice of parallel programming
Improved Grid Metascheduler Design using the Plackett-Burman Methodology
HPCS '07 Proceedings of the 21st International Symposium on High Performance Computing Systems and Applications
Proactive experiment-driven learning for system management
Proactive experiment-driven learning for system management
Future Generation Computer Systems
Active and accelerated learning of cost models for optimizing scientific applications
VLDB '06 Proceedings of the 32nd international conference on Very large data bases
| Hi-index | 0.00 |
As large-scale computational applications in various scientific domains have been utilized over many integrated sets of grid computing resources, the difficulty of their execution management and control has increased. It is beneficial to refer job history from many application executions, in order to identify application's characteristics and to decide grid resource selection policies meaningfully. In this paper, we apply a statistical technique, Plackett-Burman design with fold-over, for analyzing grid environments and execution history of applications. It identifies main factors in grid environments and applications, ranks based on how much they affect. Especially, the effective factors could be used for future resource selection. Through this process, application is performed on the selected resource and the result is added to job history. We analyzed job history from an aerospace research grid system. The effective key factors were identified and applied to resource selection policy.