论文部分内容阅读
传统Web服务器对于到达的请求使用先来先服务机制。但是,先来先服务机制无法为到达的请求提供区分服务和服务质量保证。在不精确计算和弱-硬实时调度策略基础上,提出了一种Web服务质量机制(简称WQS)为到达Web服务器的请求提供区分服务与服务质量保证。WQS不同于先来先服务机制在于:它首先利用一个请求分类器分类到达的请求到不同级别的队列。其次,借助不精确调度器与内容自适应器,WQS能够使优先级高的队列中请求具有好的响应率和短的响应延迟,使优先级低的队列中请求具有满意的响应率和可接受的响应延迟。最后,本论文描述了一系列在ApacheWeb服务器上针对.WQS的仿真实验,实验表明:WQS是一种能够提供区分服务与服务质量保证,达到改进传统Web服务器或Web服务的可行和有效的方法。
Traditional web servers use a first-come, first-served mechanism for arriving requests. However, first-come-first-serve mechanisms can not provide differentiated services and quality of service guarantees for requests arriving. Based on inaccurate computation and weak-hard real-time scheduling strategy, a Web Services Quality Mechanism (WQS) is proposed to provide differentiated services and service quality assurance for requests arriving at Web servers. WQS different from the first come first served mechanism is: It first uses a request classifier to classify the arrival of the request to a different level of the queue. Second, with inexact schedulers and content adapters, WQS enables high-priority queuing requests to have good response rates and short response latency so that requests in lower-priority queues have satisfactory response rates and are acceptable The response delay. Finally, this paper describes a series of experiments on .WQS on an Apache Web server. The experiment shows that WQS is a feasible and effective way to differentiate service and service quality assurance and to improve the traditional Web server or Web service.