Adaptive grid job scheduling with genetic algorithms

This paper proposes two models for predicting the completion time of jobs in a service Grid. The single service model predicts the completion time of a job in a Grid that provides only one type of service. The multiple services model predicts the completion time of a job that runs in a Grid which offers multiple types of services. We have developed two algorithms that use the predictive models to schedule jobs at both system level and application level. In application-level scheduling, genetic algorithms are used to minimize the average completion time of jobs through optimal job allocation on each node. The experimental results have shown that the scheduling system using the adaptive scheduling algorithms can allocate service jobs efficiently and effectively.

[1]  Rajkumar Buyya,et al.  A taxonomy and survey of grid resource management systems for distributed computing , 2002, Softw. Pract. Exp..

[2]  Hai Zhuge,et al.  China's E-Science Knowledge Grid Environment , 2004, IEEE Intell. Syst..

[3]  Chuang Liu,et al.  Design and evaluation of a resource selection framework for Grid applications , 2002, Proceedings 11th IEEE International Symposium on High Performance Distributed Computing.

[4]  Ashish Sharma,et al.  Dynamic mapping in a heterogeneous environment with tasks having priorities and multiple deadlines , 2003, Proceedings International Parallel and Distributed Processing Symposium.

[5]  Francine Berman,et al.  The AppLeS Project: A Status Report , 1997 .

[6]  Anand Sivasubramaniam,et al.  An Integrated Approach to Parallel Scheduling Using Gang-Scheduling, Backfilling, and Migration , 2001, JSSPP.

[7]  Rajesh Raman,et al.  Matchmaking: distributed resource management for high throughput computing , 1998, Proceedings. The Seventh International Symposium on High Performance Distributed Computing (Cat. No.98TB100244).

[8]  Henri Casanova,et al.  A decoupled scheduling approach for Grid application development environments , 2003, J. Parallel Distributed Comput..

[9]  Yves Robert,et al.  Optimal algorithms for scheduling divisible workloads on heterogeneous systems , 2003, Proceedings International Parallel and Distributed Processing Symposium.

[10]  Hai Zhuge.  Semantics, Resource and Grid , 2004, Future Gener. Comput. Syst..

[11]  Henri Casanova,et al.  Netsolve: a Network-Enabled Server for Solving Computational Science Problems , 1997, Int. J. High Perform. Comput. Appl..

[12]  Henri Casanova,et al.  Adaptive Scheduling for Task Farming with Grid Middleware , 1999, Int. J. High Perform. Comput. Appl..

[13]  Henri Casanova,et al.  A Modular Scheduling Approach for Grid Application Development Environments , 2002 .

[14]  David Abramson,et al.  An Evaluation of Economy-based Resource Trading and Scheduling on Computational Power Grids for Parameter Sweep Applications , 2000 .

[15]  Francine Berman,et al.  Heuristics for scheduling parameter sweep applications in grid environments , 2000, Proceedings 9th Heterogeneous Computing Workshop (HCW 2000) (Cat. No.PR00556).

[16]  Francine Berman,et al.  Logistical quality of service in NetSolve , 1999, Comput. Commun..

[17]  Miron Livny,et al.  Condor and the Grid , 2003 .

[18]  John F. Karpovich,et al.  Resource management in Legion , 1999, Future Gener. Comput. Syst..

[19]  Mark Crovella.  Performance Evaluation with Heavy Tailed Distributions , 2000, Computer Performance Evaluation / TOOLS.

[20]  Derek Wright,et al.  Cheap cycles from the desktop to the dedicated cluster: combining opportunistic and dedicated scheduling with Condor , 2007 .

[21]  Warren Smith,et al.  A Resource Management Architecture for Metacomputing Systems , 1998, JSSPP.