共查询到19条相似文献,搜索用时 500 毫秒
1.
在大型分布式仿真环境下,仿真的顺利推进取决于每一个邦员的健康运行,负载不平衡会对仿真性能造成严重的影响。作为分布式仿真事实上的标准,HLA仿真框架主要关注于解决分布式仿真的互操作和可重用问题,而没有提供对可伸缩性、动态负载平衡以及容错的支持。通过分析HLA仿真框架的不足,提出了一个基于网格服务的先进分布式仿真框架GADS,并在此基础上设计了一个负载平衡解决方案,随后重点描述了其中的负载平衡策略和基于仿真代理的高效邦员迁移算法。利用在GADS框架中引入的仿真代理层,该迁移方案大大降低了迁移开销,并实现了真正的“免冻结”迁移。 相似文献
2.
一种支持分布式进程迁移的动态负载平衡征募算法的研究 总被引:1,自引:0,他引:1
负载平衡是分布式系统必须考虑的问题,本文介绍的征募算法独立于网络拓扑结构,其思想可以应用到分布式系统中,征募算法的设计思想向传统负载平衡算法提出了挑战,它不但克服了投标算法的缺点,而且在减小通讯开销和提高处理机利用率两方面作了很多努力,使其成为一种高效的分布式进程迁移和动态负载平衡策略。我们在分布式UNIX系统上实现并验证了征募算法的高效性。 相似文献
3.
基于网格的分布式仿真系统负载平衡研究 总被引:1,自引:0,他引:1
对进程迁移的研究起源于分布式处理的出现,它使得分布式系统内动态负载平衡成为可能,并可显著提高系统的容错性和可用性;网格是新一代信息基础设施,其目的是实现资源共享利用网格提供的信息,将过载节点上的进程迁移到合适的其它节点上运行,就可以实现真正的负载平衡。该文首先简要介绍了进程迁移及网格的资源管理功能,根据现有的基于HLA构建的分布式仿真系统存在的缺陷,利用进程迁移、网格技术的优势研究并提出了分布式仿真系统负载平衡使能模块(LBEM)。现初步实现了网格环境中进程迁移的部分功能,该部分的研究工作是后续研究的基础。 相似文献
4.
随着基于HLA的分布式仿真实现技术的日益成熟,对于仿真的效率和精确性的要求也越来越高.大规模的分布式仿真中,负载平衡成为影响仿真效率和正确性的关键问题,然而HLA本身并未提供负载平衡机制.在对负载平衡问题和HLA深入研究的基础上,在HLA中实现了负载平衡机制,使用计算机性能的评价参数CPU占有率和内存使用率作为评估标准,以确定进行负载迁移的对象.最后通过实验证明,该机制能有效地平衡仿真系统中各节点之间的负载,从而验证了该机制的有效性. 相似文献
5.
数据迁移是大规模存储系统实现负载平衡的重要手段,可有效提高数据访问的性能。然而,现有数据迁移机制大多依赖于集中式算法,存在扩展性方面的不足。提出一种全分布式的数据迁移算法DBDS(Dimension By Dimension Spreading)来解决系统扩展性方面的问题。仿真实验表明,该算法在保证负载全局均衡性的同时,可有效降低迁移过程的消耗,适用于大规模分布式存储系统。 相似文献
6.
针对传统分布式仿真框架HLA/RTI中动态调度负载平衡较差、执行可靠性较低的问题,提出一种基于环境感知的仿真服务调度模型C3SM。C3SM模型包含总体结构、调度策略和服务部署。总体结构规定各个模块的功能与交互接口,调度策略采用改进的蚁群算法进行仿真服务的动态调度,服务部署采用重叠覆盖部署策略以保证较好的服务可用性和较低的资源消耗。仿真实验结果表明,与传统分布式仿真框架HLA/RTI相比,C3SM模型通过实时获取执行环境的上下文信息,能够获得较优的负载平衡度,且重叠覆盖的仿真服务部署策略使得仿真运行系统的可靠性明显提高。 相似文献
7.
8.
在分布式 列系统中保持全系统的负载平衡是非常重要的一个环节。进程动态迁移可在一种透明的工作方式下改善系统的性能。本文对分布式磁盘阵列系统中进程的动态迁移进行了系统的分析,并提出一优化算法,在一定程度上增强了现有系统的性能。 相似文献
9.
10.
基于网格的两级动态负载平衡算法 总被引:1,自引:1,他引:0
网格系统具有异构性、动态性和分布性的特点,且资源数量巨大,这使得网格中的任务调度十分复杂.针对网格的特点,在两级树型网格结构的基础上,设计了一种基于该结构的两级负载平衡算法,针对传统的负载平衡算法考虑资源因素单一,难以满足复杂的网格环境的要求,该算法使用了多种负载参数来衡量网格节点的负载状况. 相似文献
11.
Load balancing has been a key concern for traditional multiprocessor systems. The emergence of computational grids extends this challenge to deal with more serious problems, such as scalability, heterogeneity of computing resources and considerable transfer delay. In this paper, we present a dynamic and decentralized load balancing algorithm for computationally intensive jobs on a heterogeneous distributed computing platform. The time spent by a job in the system is considered as the main issue that needs to be minimized. Our main contributions are: (1) Our algorithm uses site desirability for processing power and transfer delay to guide load assignment and redistribution, (2) Our transfer and location policies are a combination of two specific strategies that are performance driven to minimize execution cost. These two policies are the Instantaneous Distribution Policy (IDP) and the Load Adjustment Policy (LAP), (3) The communication overhead involved in information collection is reduced using mutual information feedback. The simulation results show that our proposed algorithm outperforms conventional approaches over a wide range of system parameters. 相似文献
12.
13.
Yongsheng HaoAuthor Vitae Guanfeng LiuAuthor VitaeNa WenAuthor Vitae 《Future Generation Computer Systems》2012,28(4):657-665
A Grid is a network of computational resources that may potentially span many continents. Load balancing in a Grid is a hot research issue which affects every aspect of the Grid, including service selection and task execution. Thus, it is necessary and significant to solve the load balancing problem in a Grid. In this paper, we propose a dynamic, distributed load balancing scheme for a Grid which provides deadline control for tasks. In our scenario, first, resources check their state and make a request to the Grid Broker according to the change of load state. Then, the Grid Broker assigns Gridlets between resources and scheduling for load balancing under the deadline request. We apply our load balancing strategy into a popular Grid simulation platform GridSim. Experimental results prove that our proposed load balancing mechanism can (1) reduce the makespan, (2) improve the finished rate of the Gridlet, and (3) reduce the resubmitted time. 相似文献
14.
15.
16.
17.
基于规则的分层负载平衡调度模型 总被引:13,自引:0,他引:13
On a massively parallel and distributed system and a network of workstations system, it is a critical problem to increase the utilization efficiency of resources and the answer speed of tasks by using effective load balancing scheduling strategy. This paper analyzes the scheduling strategy of dynamic load balancing and static load balancing,and then proposes a hierarchical load balancing scheduling model based on rules. Finally,making somecomparisons with Other scheduling models. 相似文献
18.
19.
The paper describes methods for using Extremal Optimization (EO) for processor load balancing during execution of distributed applications. A load balancing algorithm for clusters of multicore processors is presented and discussed. In this algorithm the EO approach is used to periodically detect the best tasks as candidates for migration and for a guided selection of the best computing nodes to receive the migrating tasks. To decrease the complexity of selection for migration, the embedded EO algorithm assumes a two-step stochastic selection during the solution improvement based on two separate fitness functions. The functions are based on specific models which estimate relations between the programs and the executive hardware. The proposed load balancing algorithm is assessed by experiments with simulated load balancing of distributed program graphs. The algorithm is compared against a greedy fully deterministic approach, a genetic algorithm and an EO-based algorithm with random placement of migrated tasks. 相似文献