摘要： 由于客户机访问服务器的随机性，使得各中间应用服务器所接收到的任务出现差异，在某些时刻，可能有的中间应用服务器负载过重，而同时有的中间应用服务器却很空闲或完全闲置．此时，为充分利用网络硬件资源及缩短客户等待时间，可将负载过重的中间应用服务器的一部分任务转移给较空闲的中间应用服务器，从而使各中间应用服务器的负载大致均衡．为此，该文通过引入两个物理模型及相应算法解决这一问题，并为解决类似问题开辟了一条新的途径．

关键词: 中间应用服务器, 任务均衡, 水压模型, 水库模型

Abstract: Because of the randomicity when client calls server, there is difference between tasks to be received by every middle application server. sometime, some application servers will load to much while some ones else will be very idle or even no task. In order to fully utilizing the network hardware resources and shortening the clients' waiting time, we can transfer a part of tasks of application server whose loads are too heavy to those which are idler it is a good means that every application server's load will be approximate balance. This article uses two physical models and homologous arithmetic to solve this problem and inaugurates a new approach which can be used to solve similar problems.

中图分类号: 

• TP311.52