论文部分内容阅读
为了提高Web服务器处理用户请求的能力,大多数Web服务器已经采用多核处理器.但传统的动态请求调度算法,例如先到先服务算法,既没有考虑多核处理器的结构特征也没有考虑动态请求服务时间的分布,因此不能充分利用多核处理器的运算能力.为解决此问题,在之前工作的基础上提出一种新的动态请求调度算法.新算法可周期性地预测表征负载状态的动态请求队列的权值,根据预测的权值确定动态请求分配方案,并依据此方案对到达Web服务器的动态请求进行调度.为评估此算法的性能,以此算法为基础进行了仿真实验.实验结果表明,新的算法不仅能解决多核Web服务器中存在的“乒乓”效应等影响性能发挥的问题,而且能在较长时间内保持处理器核心间的负载均衡.
Most Web servers have adopted multi-core processors in order to improve the ability of Web servers to process user requests.However, the traditional dynamic request scheduling algorithms, such as the first-come-first-serve algorithm, neither consider the architectural features of multi-core processors nor consider the dynamic request service So we can not take full advantage of the computing power of multi-core processors.In order to solve this problem, a new dynamic request scheduling algorithm is proposed based on the previous work.The new algorithm can periodically predict the dynamic request queue The dynamic request allocation scheme is determined according to the predicted weight, and the dynamic request dispatching to the Web server is scheduled according to the scheme.In order to evaluate the performance of this algorithm, a simulation experiment is carried out based on the algorithm.Experimental results show that, The new algorithm can not only solve the problem of performance such as “ping-pong” effect existing in multi-core Web server, but also keep load balance among processor cores for a long time.