Architectural Characterization of Processor Affinity in Network Processing

IEEE International Symposium on Performance Analysis of Systems and Software, 2005. ISPASS 2005. Pub Date : 2005-03-20 DOI:10.1109/ISPASS.2005.1430575

A. Foong, Jason M. Fung, D. Newell, S. Abraham, Peggy Irelan, Alex A. Lopez-Estrada

{"title":"Architectural Characterization of Processor Affinity in Network Processing","authors":"A. Foong, Jason M. Fung, D. Newell, S. Abraham, Peggy Irelan, Alex A. Lopez-Estrada","doi":"10.1109/ISPASS.2005.1430575","DOIUrl":null,"url":null,"abstract":"Network protocol stacks, in particular TCP/IP software implementations, are known for its inability to scale well in general-purpose monolithic operating systems (OS) for SMP. Previous researchers have experimented with affinitizing processes/thread, as well as interrupts from devices, to specific processors in a SMP system. However, general purpose operating systems have minimal consideration of user-defined affinity in their schedulers. Our goal is to expose the full potential of affinity by in-depth characterization of the reasons behind performance gains. We conducted an experimental study of TCP performance under various affinity modes on IA-based servers. Results showed that interrupt affinity alone provided a throughput gain of up to 25%, and combined thread/process and interrupt affinity can achieve gains of 30%. In particular, calling out the impact of affinity on machine clears (in addition to cache misses) is characterization that has not been done before","PeriodicalId":230669,"journal":{"name":"IEEE International Symposium on Performance Analysis of Systems and Software, 2005. ISPASS 2005.","volume":"55 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2005-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"38","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE International Symposium on Performance Analysis of Systems and Software, 2005. ISPASS 2005.","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISPASS.2005.1430575","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 38

Abstract

Network protocol stacks, in particular TCP/IP software implementations, are known for its inability to scale well in general-purpose monolithic operating systems (OS) for SMP. Previous researchers have experimented with affinitizing processes/thread, as well as interrupts from devices, to specific processors in a SMP system. However, general purpose operating systems have minimal consideration of user-defined affinity in their schedulers. Our goal is to expose the full potential of affinity by in-depth characterization of the reasons behind performance gains. We conducted an experimental study of TCP performance under various affinity modes on IA-based servers. Results showed that interrupt affinity alone provided a throughput gain of up to 25%, and combined thread/process and interrupt affinity can achieve gains of 30%. In particular, calling out the impact of affinity on machine clears (in addition to cache misses) is characterization that has not been done before

查看原文本刊更多论文

网络处理中处理器亲和性的体系结构表征

众所周知，网络协议栈，特别是TCP/IP软件实现，无法很好地扩展到用于SMP的通用单片操作系统(OS)中。先前的研究人员已经尝试将进程/线程以及来自设备的中断关联到SMP系统中的特定处理器。但是，通用操作系统在其调度器中很少考虑用户定义的关联。我们的目标是通过深入描述性能提升背后的原因，揭示亲和力的全部潜力。我们在基于ia的服务器上对不同亲和模式下的TCP性能进行了实验研究。结果表明，中断亲和性单独提供了高达25%的吞吐量增益，而线程/进程和中断亲和性相结合可以实现30%的增益。特别是，指出亲和性对机器清除(除了缓存丢失)的影响是以前从未做过的特性描述

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

IEEE International Symposium on Performance Analysis of Systems and Software, 2005. ISPASS 2005.

自引率

0.00%

发文量