首页> 外文会议>Distributed Computing >On the Emulation of Finite-Buffered Output Queued Switches Using Combined Input-Output Queuing
【24h】

On the Emulation of Finite-Buffered Output Queued Switches Using Combined Input-Output Queuing

机译:基于组合输入输出排队的有限缓冲输出排队开关仿真

获取原文
获取原文并翻译 | 示例

摘要

We study the emulation of Output Queuing (OQ) using Combined Input-Output Queuing (CIOQ) switches in the setting where the emulated OQ switch and the CIOQ switch have buffer capacity B ≥ 1 packets at every output. We analyze the resource requirements of CIOQ policies in terms of the required fabric speedup and the additional buffer capacity needed at the CIOQ inputs. For the family of work-conserving scheduling algorithms, we find that whereas every greedy CIOQ policy is valid for OQ emulation at speedup B, no CIOQ policy is valid at speedup s < B-2~(1/3) when preemption is allowed. We also find that CCF in particular is not valid at any speedup s < B. We then introduce a CIOQ policy, CEH, that is valid at speedup s ≥ 2(B-1)~(1/2). Under CEH, the buffer occupancy at any input never exceeds 1 + [B-1/s-1]. For non-preemptive scheduling algorithms, we characterize a trade-off between the CIOQ speedup and the input buffer occupancy. Specifically, we show that for any greedy policy that is valid at speedup s > 2, the input buffer occupancy cannot exceed 1 + [B-1/s-2]. We also show that at speedup 2, a greedy variant of the CCF policy requires input buffer capacity of only B packets for the emulation of non-preemptive OQ algorithms with PIFO service disciplines.
机译:我们在模拟OQ开关和CIOQ开关在每个输出的缓冲区容量B≥1数据包的设置中研究了使用组合输入输出排队(CIOQ)开关的输出排队(OQ)的仿真。我们根据所需的架构加速和CIOQ输入所需的额外缓冲区容量来分析CIOQ策略的资源需求。对于节省工作量的调度算法系列,我们发现,虽然每个贪婪的CIOQ策略在加速B时都适用于OQ仿真,但在允许先占的情况下,在加速s 2时有效的任何贪心策略,输入缓冲区占用率不能超过1 + [B-1 / s-2]。我们还显示,在速度提高2时,CCF策略的贪婪变体仅需要B数据包的输入缓冲区容量,即可使用PIFO服务规范来模拟非抢先OQ算法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号