基于网格计算的自适应负载均衡策略研究

The Research of Auto-adaptive Load Balancing Policy Based on Grid Computing

为了融合大量网络资源并有效地计算,解决网格计算中的负载均衡问题成为关键性的技术。论文提出了一种自适应负载均衡策略,采用了以静态为辅,动态自适应负载均衡算法为主的服务,可根据具体计算任务的情况,对任务重定向分配,提高了系统的伸缩性和响应时间,并采用基于CORBA体系机构的设计,在网格中间件层服务,灵活选择负载均衡算法,达到系统透明性。通过仿真模拟,证实了此策略的实用性和有效性。

Load balancing in the grid computing environments is the key technology by way of combining the broad-field internet resource. The paper has put forward an auto-adaptive load balancing policy. The policy is used the primary of dynamic adaptive load balancing algorithm assisted with static load balancing, According to the computing task on detail, improved the retraetility and responding time of system by redistributing the subtask, and adopted the design of CORBA-based system, enabled the load balancing system transparent with the grid middleware services choose agilely load balancing algorithms and improved the efficiency of computing. The paper has approved the practicability and validity of the load balancing policy by the laboratorial simulation.