TCP Adaptation for MPI on Long-and-Fat Networks
TCP Adaptation for MPI on Long-and-Fat Networks
复制标题
长而宽的网络上 MPI 的 TCP 适配
DOI：
10.1109/clustr.2005.347034
复制
发表时间：
2005
期刊：
影响因子：
--
通讯作者：
Y. Ishikawa
中科院分区：
文献类型：
--
作者：
Motohiko Matsuda;T. Kudoh;Yuetsu Kodama;Ryousei Takano;Y. Ishikawa
典型的MPI应用程序在计算和通信阶段工作，消息以相对较小的块进行交换。此行为不适合于TCP，因为TCP仅设计用于高效地处理连续的消息流。这种行为异常是众所周知的，但修复程序没有集成到今天的TCP实现中，即使性能严重下降，特别是对于MPI应用程序。本文对Linux的TCP协议栈提出了三点改进：启动时的调步、减少重传超时时间、在MPI应用程序的计算阶段转换时切换TCP参数。使用NAS并行基准对这些改进进行的评估显示，BT、CG、IS和SP基准实现了10%到30%的改进。另一方面，FT和MG基准测试没有显示出任何改进，因为它们具有TCP假定的稳定通信，而LU基准测试稍微变差了，因为它的通信非常少
Typical MPI applications work in phases of computation and communication, and messages are exchanged in relatively small chunks. This behavior is not optimal for TCP because TCP is designed only to handle a contiguous flow of messages efficiently. This behavior anomaly is well-known, but fixes are not integrated into today's TCP implementations, even though performance is seriously degraded, especially for MPI applications. This paper proposes three improvements in the Linux TCP stack: i.e., pacing at start-up, reducing Retransmit-Timeout time, and TCP parameter switching at the transition of computation phases in an MPI application. Evaluation of these improvements using the NAS parallel benchmarks shows that the BT, CG, IS, and SP benchmarks achieved 10 to 30 percent improvements. On the other hand, the FT and MG benchmarks showed no improvement because they have the steady communication that TCP assumes, and the LU benchmark became slightly worse because it has very little communication
长管网络上并行 TCP 流的层间协调
DOI：
--
发表时间：
2004
期刊：
SC2004 (Web)
影响因子：
--
作者：
Hiroyuki Kamezawa;Makoto Nakamura;Mary Inaba;Kei Hiraki 他
通讯作者：
Kei Hiraki 他