{"title":"以轮循平摊代价提供近乎最优的公平排队保证","authors":"P. Valente","doi":"10.1109/ICCCN.2013.6614175","DOIUrl":null,"url":null,"abstract":"Round-robin schedulers are the most efficient solution for providing strong QoS guarantees on high-speed links. Yet these schedulers suffer from a high worst-case delay with respect to an ideal, perfectly fair service. More costly schedulers are needed to provide better service guarantees. In this paper we tackle this problem by proposing a simple modification scheme for reducing the amortized execution time of fair-queueing schedulers. We prove that, applying this scheme to existing accurate schedulers, we can define new schedulers providing near-optimal service guarantees at an amortized computational cost close to that of just Deficit Round Robin (DRR). Finally, we show Quick Fair Queueing Plus (QFQ+), a new scheduler obtained by applying our scheme to QFQ. QFQ+ replaced QFQ in the Linux kernel. According to our experimental results, and exactly in the scenarios where QFQ+ provides better service guarantees than a round-robin scheduler, the time and the energy needed to process packets with QFQ+ is lower than with DRR.","PeriodicalId":207337,"journal":{"name":"2013 22nd International Conference on Computer Communication and Networks (ICCCN)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2013-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":"{\"title\":\"Providing Near-Optimal Fair-Queueing Guarantees at Round-Robin Amortized Cost\",\"authors\":\"P. Valente\",\"doi\":\"10.1109/ICCCN.2013.6614175\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Round-robin schedulers are the most efficient solution for providing strong QoS guarantees on high-speed links. Yet these schedulers suffer from a high worst-case delay with respect to an ideal, perfectly fair service. More costly schedulers are needed to provide better service guarantees. In this paper we tackle this problem by proposing a simple modification scheme for reducing the amortized execution time of fair-queueing schedulers. We prove that, applying this scheme to existing accurate schedulers, we can define new schedulers providing near-optimal service guarantees at an amortized computational cost close to that of just Deficit Round Robin (DRR). Finally, we show Quick Fair Queueing Plus (QFQ+), a new scheduler obtained by applying our scheme to QFQ. QFQ+ replaced QFQ in the Linux kernel. According to our experimental results, and exactly in the scenarios where QFQ+ provides better service guarantees than a round-robin scheduler, the time and the energy needed to process packets with QFQ+ is lower than with DRR.\",\"PeriodicalId\":207337,\"journal\":{\"name\":\"2013 22nd International Conference on Computer Communication and Networks (ICCCN)\",\"volume\":\"1 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2013-10-24\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"6\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2013 22nd International Conference on Computer Communication and Networks (ICCCN)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICCCN.2013.6614175\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2013 22nd International Conference on Computer Communication and Networks (ICCCN)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCCN.2013.6614175","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 6
摘要
循环调度程序是在高速链路上提供强QoS保证的最有效的解决方案。然而,与理想的、完全公平的服务相比,这些调度器遭受了很高的最坏情况延迟。需要更昂贵的调度器来提供更好的服务保证。本文通过提出一种简单的修改方案来减少公平排队调度程序的平摊执行时间,从而解决了这个问题。我们证明,将此方案应用于现有的精确调度器,我们可以定义新的调度器,以接近赤字轮询(DRR)的平摊计算成本提供接近最优的服务保证。最后,我们展示了将我们的方案应用到QFQ上得到的一个新的调度程序Quick Fair Queueing Plus (QFQ+)。QFQ+取代了Linux内核中的QFQ。根据我们的实验结果,并且正是在QFQ+提供比轮循调度程序更好的服务保证的场景下,使用QFQ+处理数据包所需的时间和能量低于DRR。
Providing Near-Optimal Fair-Queueing Guarantees at Round-Robin Amortized Cost
Round-robin schedulers are the most efficient solution for providing strong QoS guarantees on high-speed links. Yet these schedulers suffer from a high worst-case delay with respect to an ideal, perfectly fair service. More costly schedulers are needed to provide better service guarantees. In this paper we tackle this problem by proposing a simple modification scheme for reducing the amortized execution time of fair-queueing schedulers. We prove that, applying this scheme to existing accurate schedulers, we can define new schedulers providing near-optimal service guarantees at an amortized computational cost close to that of just Deficit Round Robin (DRR). Finally, we show Quick Fair Queueing Plus (QFQ+), a new scheduler obtained by applying our scheme to QFQ. QFQ+ replaced QFQ in the Linux kernel. According to our experimental results, and exactly in the scenarios where QFQ+ provides better service guarantees than a round-robin scheduler, the time and the energy needed to process packets with QFQ+ is lower than with DRR.