This paper suggests a hybrid resource management approach for efficient paralleldistributed computing on the Grid. It operates on both application and system levels, combining user-level job scheduling with dynamic w...
详细信息
This paper suggests a hybrid resource management approach for efficient paralleldistributed computing on the Grid. It operates on both application and system levels, combining user-level job scheduling with dynamic workload balancing algorithm that automatically adapts a parallelapplication to the heterogeneous resources, based on the actual resource parameters and estimated requirements of the application. The hybrid environment and the algorithm for automated load balancing are described, the influence of resource heterogeneity level is measured, and the speedup achieved with this technique is demonstrated for different types of applications and resources. (c) 2008 Elsevier B.V. All rights reserved.
In this paper we address the critical issues of efficient resource management and high-performance paralleldistributed computing on the Grid by introducing a new hierarchical approach that combines a user-level job s...
详细信息
ISBN:
(纸本)9783540739395
In this paper we address the critical issues of efficient resource management and high-performance paralleldistributed computing on the Grid by introducing a new hierarchical approach that combines a user-level job scheduling with a dynamic load balancing technique that automatically adapts a black-box distributed or parallelapplication to the heterogeneous resources. The algorithm developed dynamically selects the resources best suited for a particular task or parallel process of the executed application, and optimizes the load balance based on the dynamically measured resource parameters and estimated requirements of the application. We describe the proposed algorithm for automated load balancing, paying attention to the influence of resource heterogeneity metrics, demonstrate the speedup achieved with this technique for different types of applications and resources, and propose a way to extend the approach to a wider class of applications.
This paper addresses a problem of porting a distributedparallelapplication to the Grid. As a case study we use the Virtual Reactor application on the Russian-Dutch Grid testbed. We sketch the Grid testbed infrastruc...
详细信息
ISBN:
(纸本)3540343792
This paper addresses a problem of porting a distributedparallelapplication to the Grid. As a case study we use the Virtual Reactor application on the Russian-Dutch Grid testbed. We sketch the Grid testbed infrastructure and application modular architecture, and concentrate on performance issues of one of the core parallel solvers on the Grid. We compare the performance achieved on homogeneous resources with that observed on heterogeneous computing and networking infrastructure. To increase the parallel efficiency of the solver on heterogeneous resources we developed an adaptive load balancing algorithm. We demonstrate the speedup achieved with this technique and indicate the ways to further enhance the algorithm and develop an automated procedure for optimal utilization of Grid resources for parallel computing.
暂无评论