Fairness scheduler for multithreaded programs in Virtual Machine environment

S. Mostafa, S. Kusakabe, H. Amano
{"title":"Fairness scheduler for multithreaded programs in Virtual Machine environment","authors":"S. Mostafa, S. Kusakabe, H. Amano","doi":"10.1109/JEC-ECC.2016.7518972","DOIUrl":null,"url":null,"abstract":"Fairness measure is used in designing Operating System (OS) schedulers to determine whether users or applications are receiving a fair share of system resources. PS (proportional share) schedulers are the dominant schedulers in Virtual Machine (VM) environment. A PS scheduler assigns a weight to every VM to allocate CPU cycles, allocating these cycles to VM is proportional to that weight. Current VM scheduling algorithms don't aware about the attributes (e.g. weights and number of threads) of the running processes in VMs. We illustrate this issue through experimental evaluations of the current scheduling thus revealing its weakness. A modification of current scheduler is proposed in this work to enable users to specify weights to VMs. Therefore, user gets CPU usage proportion to the assigned weight, the higher weight the more CPU usage and vice versa. In addition, a pricing calculator is considered to calculate the fee that the user should pay according to his desired CPU usage. Derived results show that our modification achieves accurate proportional fairness for a diverse set of workloads.","PeriodicalId":362288,"journal":{"name":"2016 Fourth International Japan-Egypt Conference on Electronics, Communications and Computers (JEC-ECC)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2016-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2016 Fourth International Japan-Egypt Conference on Electronics, Communications and Computers (JEC-ECC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/JEC-ECC.2016.7518972","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Fairness measure is used in designing Operating System (OS) schedulers to determine whether users or applications are receiving a fair share of system resources. PS (proportional share) schedulers are the dominant schedulers in Virtual Machine (VM) environment. A PS scheduler assigns a weight to every VM to allocate CPU cycles, allocating these cycles to VM is proportional to that weight. Current VM scheduling algorithms don't aware about the attributes (e.g. weights and number of threads) of the running processes in VMs. We illustrate this issue through experimental evaluations of the current scheduling thus revealing its weakness. A modification of current scheduler is proposed in this work to enable users to specify weights to VMs. Therefore, user gets CPU usage proportion to the assigned weight, the higher weight the more CPU usage and vice versa. In addition, a pricing calculator is considered to calculate the fee that the user should pay according to his desired CPU usage. Derived results show that our modification achieves accurate proportional fairness for a diverse set of workloads.
虚拟机环境下多线程程序的公平性调度器
公平性度量用于设计操作系统(OS)调度器,以确定用户或应用程序是否获得了公平的系统资源份额。PS (proportional share)调度器是虚拟机(VM)环境中的主流调度器。PS调度器为每个VM分配权重来分配CPU周期,将这些周期分配给VM是与该权重成比例的。当前的虚拟机调度算法不知道虚拟机中运行进程的属性(例如权重和线程数)。我们通过对当前调度的实验评估来说明这个问题,从而揭示了它的弱点。本文提出了对当前调度程序的修改,使用户能够指定vm的权重。因此,用户获得的CPU使用率与分配的权重的比例,权重越大,CPU使用率越高,反之亦然。此外,还考虑使用定价计算器来计算用户应根据其所需的CPU使用情况支付的费用。导出的结果表明,我们的修改对于不同的工作负载集实现了准确的比例公平性。
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
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信