ITR: Improving the Performance of Web Servers under Overload

这个建议的目的是在静态和动态内容过载的情况下，提供对服务器行为的深刻理解。该方法是跨学科的：有一个理论组件，旨在证明具有时间依赖负载的M/G/1队列的新定理；一个实现组件，它对web服务器和数据库服务器在过载情况下的性能作为各种网络参数的函数进行严格的系统研究；以及一个工作负载和客户端特征组件，旨在了解瞬态过载期间请求流的统计属性，并将这些属性合并到过载的web工作负载生成器的设计中。这个建议还引入了一个非常简单的想法来提高web服务器在过载情况下的性能：基于内核的SRPT连接调度实现。这个想法简单而直观。连接调度不需要购买更多的硬件，或任何其他昂贵的系统升级。它似乎允许服务器在更长的过载期间存活下来，并允许客户端在这些过载期间体验更好的性能。此外，它似乎使服务器能够更快地从过载中恢复过来。该提案将详细介绍SRPT连接调度的好处，并旨在为数据库引入类似的调度策略。以上总结了该建议的智力价值部分。提议的活动产生的更广泛的影响包括项目期间生成的所有源代码的在线可用性，包括对Apache代码和Linux的修改，以及项目期间收集的所有跟踪日志的在线可用性，以及用于过载的web工作负载生成器。该项目启动了多个学科之间的合作，包括：数学、计算机系统和运营管理，以及大学和工业界之间的合作。该项目还将为少数民族本科生提供研究机会。该方案涉及1名教师（仅1个夏季月），1.5名研究生和1名本科生，为期3年。

This proposal aims to provide this deep understanding of server behavior during overload, in thecase of both static and dynamic content. The approach is interdisciplinary: There is a theoreticalcomponent which aims to prove new theorems on M/G/1 queues with time-dependent load; an implementationcomponent which performs a rigorous systematic study of web server and database serverperformance under overload as a function of various networking parameters; and a workload and clientcharacterization component which aims to understand statistical properties of the request stream duringtransient overload and incorporate these into the design of a web workload generator for overload.This proposal also introduces a very simple idea for improving the performance of web serversunder overload: a kernel-based implementation of SRPT connection scheduling. The idea is simpleand intuitive. Connection scheduling does not require purchasing more hardware, or any other costlysystem upgrades. It appears to allow the server to survive much longer periods of overload, and to allowclients to experience better performance during these periods of overload. Furthermore, it appearsto enable the server to recuperate from overload much faster. The proposal will provide a detailedevaluation of the benefits of SRPT connection scheduling and will aim to introduce similar schedulingpolicies for databases. The above concludes the intellectual merit portion of the proposal.The broader impacts resulting from the proposed activity include online availability of all sourcecode generated during the project including modifications to the Apache code and to Linux, as wellas online availability of all trace logs collected during this project, and a web workload generator foroverload. The project initiates collaborations across various academic disciplines including: mathematics,computer systems, and operations management, as well as between universities and industry.The project will also provide a research opportunity for a minority undergraduate student.The proposal involves 1 faculty member (1 summer month only), 1.5 graduate students, and 1undergraduate, over the course of 3 years.