网上交易系统性能优化算法

基于TPC-W基准对网上交易系统的性能测试及分析进行研究,找出了影响系统性能的瓶颈,进而提出一种基于TPC-W的网上交易系统性能优化算法.该算法使用队列调度和管理来协助实现拥塞控制,根据先到先服务的原则,公平处理到达缓冲区的服务请求,能有效解决系统拥塞问题,从而达到算法优化,提高系统性能的目的.通过实验可知,使用优化算法前,交易响应时间随着远程浏览器(RBE)数量增加而迅速提高;而使用优化算法之后,每种交易的响应时间提高到一定程度之后趋于稳定,符合优化算法设计的预期效果.

Optimization algorithm for Web trade system

According to the TPC-W(transaction processing performance concil-Web e-commerce)benchmark,a TPC-W performance test and analyses of the trading system is presented.The bottleneck of the performance was found.And an optimization algorithm of the Web trading system based on the TPC-W benchmark is introduced.The algorithm controls all of congestion by assistance of alignment attemptering and managing,it redresses the balance all of service requests that arrive at the cache according a special principia.The special principia is like this: first arrive,first servicing.That method could solve the congestion efficiently.So optimization algorithm can improve performance of the whole system.By contrasting the two experiment results which before and after using the optimization algorithm,it can be easily found that if we use the optimization algorithm,the response time of each transaction will trend to be stable.Before using the optimization algorithm,the transaction response time will become longer and longer until the server can not give any response.Then it can be concluded that this experiment result accords with the purpose of the optimization algorithm.