Improvement of a TCP Incast avoidance method for data center networks

Kazutoshi Kajita, Shigeyuki Osada, Yukinobu Fukushima, Tokumi Yokohira

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

In distributed file systems, a well-known congestion collapse called TCP Incast occurs because many servers send data to the same client and then many packets overflow the port buffer of the client link. Incast leads to throughput degradation in the network. In our previous work, we have proposed a method to avoid Incast. In the method, we limit the maximum number of simultaneously existing connections to a predetermined constant value. However, we cannot use the method when the data size is small, and because we have not investigated how to optimize the maximum value, Incast may occur if the maximum value is not appropriate. In this paper, we improve the method so that it is applicable regardless of the data size and the maximum value is optimized. Numerical results show the effectiveness of our proposed method.

Original languageEnglish
Title of host publicationInternational Conference on ICT Convergence
PublisherIEEE Computer Society
Pages459-464
Number of pages6
ISBN (Print)9781479906987
DOIs
Publication statusPublished - 2013
Event2013 International Conference on Information and Communication Technology Convergence, ICTC 2013 - Jeju Island, Korea, Republic of
Duration: Oct 14 2013Oct 16 2013

Other

Other2013 International Conference on Information and Communication Technology Convergence, ICTC 2013
CountryKorea, Republic of
CityJeju Island
Period10/14/1310/16/13

Fingerprint

Servers
Throughput
Degradation

Keywords

  • Data Center
  • Distributed File System
  • TCP
  • TCP Incast

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems

Cite this

Kajita, K., Osada, S., Fukushima, Y., & Yokohira, T. (2013). Improvement of a TCP Incast avoidance method for data center networks. In International Conference on ICT Convergence (pp. 459-464). [6675396] IEEE Computer Society. https://doi.org/10.1109/ICTC.2013.6675396

Improvement of a TCP Incast avoidance method for data center networks. / Kajita, Kazutoshi; Osada, Shigeyuki; Fukushima, Yukinobu; Yokohira, Tokumi.

International Conference on ICT Convergence. IEEE Computer Society, 2013. p. 459-464 6675396.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kajita, K, Osada, S, Fukushima, Y & Yokohira, T 2013, Improvement of a TCP Incast avoidance method for data center networks. in International Conference on ICT Convergence., 6675396, IEEE Computer Society, pp. 459-464, 2013 International Conference on Information and Communication Technology Convergence, ICTC 2013, Jeju Island, Korea, Republic of, 10/14/13. https://doi.org/10.1109/ICTC.2013.6675396
Kajita K, Osada S, Fukushima Y, Yokohira T. Improvement of a TCP Incast avoidance method for data center networks. In International Conference on ICT Convergence. IEEE Computer Society. 2013. p. 459-464. 6675396 https://doi.org/10.1109/ICTC.2013.6675396
Kajita, Kazutoshi ; Osada, Shigeyuki ; Fukushima, Yukinobu ; Yokohira, Tokumi. / Improvement of a TCP Incast avoidance method for data center networks. International Conference on ICT Convergence. IEEE Computer Society, 2013. pp. 459-464
@inproceedings{7cc917c4d62c4c128425b218256234b4,
title = "Improvement of a TCP Incast avoidance method for data center networks",
abstract = "In distributed file systems, a well-known congestion collapse called TCP Incast occurs because many servers send data to the same client and then many packets overflow the port buffer of the client link. Incast leads to throughput degradation in the network. In our previous work, we have proposed a method to avoid Incast. In the method, we limit the maximum number of simultaneously existing connections to a predetermined constant value. However, we cannot use the method when the data size is small, and because we have not investigated how to optimize the maximum value, Incast may occur if the maximum value is not appropriate. In this paper, we improve the method so that it is applicable regardless of the data size and the maximum value is optimized. Numerical results show the effectiveness of our proposed method.",
keywords = "Data Center, Distributed File System, TCP, TCP Incast",
author = "Kazutoshi Kajita and Shigeyuki Osada and Yukinobu Fukushima and Tokumi Yokohira",
year = "2013",
doi = "10.1109/ICTC.2013.6675396",
language = "English",
isbn = "9781479906987",
pages = "459--464",
booktitle = "International Conference on ICT Convergence",
publisher = "IEEE Computer Society",
address = "United States",

}

TY - GEN

T1 - Improvement of a TCP Incast avoidance method for data center networks

AU - Kajita, Kazutoshi

AU - Osada, Shigeyuki

AU - Fukushima, Yukinobu

AU - Yokohira, Tokumi

PY - 2013

Y1 - 2013

N2 - In distributed file systems, a well-known congestion collapse called TCP Incast occurs because many servers send data to the same client and then many packets overflow the port buffer of the client link. Incast leads to throughput degradation in the network. In our previous work, we have proposed a method to avoid Incast. In the method, we limit the maximum number of simultaneously existing connections to a predetermined constant value. However, we cannot use the method when the data size is small, and because we have not investigated how to optimize the maximum value, Incast may occur if the maximum value is not appropriate. In this paper, we improve the method so that it is applicable regardless of the data size and the maximum value is optimized. Numerical results show the effectiveness of our proposed method.

AB - In distributed file systems, a well-known congestion collapse called TCP Incast occurs because many servers send data to the same client and then many packets overflow the port buffer of the client link. Incast leads to throughput degradation in the network. In our previous work, we have proposed a method to avoid Incast. In the method, we limit the maximum number of simultaneously existing connections to a predetermined constant value. However, we cannot use the method when the data size is small, and because we have not investigated how to optimize the maximum value, Incast may occur if the maximum value is not appropriate. In this paper, we improve the method so that it is applicable regardless of the data size and the maximum value is optimized. Numerical results show the effectiveness of our proposed method.

KW - Data Center

KW - Distributed File System

KW - TCP

KW - TCP Incast

UR - http://www.scopus.com/inward/record.url?scp=84899461584&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84899461584&partnerID=8YFLogxK

U2 - 10.1109/ICTC.2013.6675396

DO - 10.1109/ICTC.2013.6675396

M3 - Conference contribution

SN - 9781479906987

SP - 459

EP - 464

BT - International Conference on ICT Convergence

PB - IEEE Computer Society

ER -