AgileDCN: An Agile Reconfigurable Optical Data Center Network Architecture

Dinh Danh Le; Liam P. Barry; Daniel C. Kilper; Philip Perry; Jingyan Wang; Conor McArdle

doi:10.1109/JLT.2020.2999778

AgileDCN: An Agile Reconfigurable Optical Data Center Network Architecture

Dinh Danh Le, Liam P. Barry, Daniel C. Kilper, Philip Perry, Jingyan Wang, Conor McArdle

Research output: Contribution to journal › Article › peer-review

3 Citations (Scopus)

86 Downloads (Pure)

Abstract

This paper presents a detailed examination of a novel data center network (DCN) that can satisfy the high capacity and low latency requirements of modern cloud computing applications. This reconfigurable architecture called AgileDCN uses fast-switching optical components with a centralized control function and workload scheduler. By providing a highly flexible optical network fabric between server racks, very high network efficiencies can be achieved even under imbalanced loading patterns. Our simulation results show that, at high (70%) loads, TCP flow completion times in the AgileDCN are significantly lower than in an equivalent electronic leaf-spine network.

Original language	English
Article number	9107450
Pages (from-to)	4922-4934
Number of pages	13
Journal	Journal of Lightwave Technology
Volume	38
Issue number	18
Early online date	3 Jun 2020
DOIs	https://doi.org/10.1109/JLT.2020.2999778
Publication status	Published - 15 Sep 2020

Bibliographical note

Funding Information:
This work was supported in part by the National Science Foundation (NSF) under Grant CNS-1827923, Grant CNS-1650669, Grant CNS-1737453, and Grant EEC-0812072, and in part by the Science Foundation of Ireland (SFI) under the US-Ireland R&D Partnership Programme (SFI project code 15/USC2C/I3132), the SFI Centres CONNECT, and IPIC.
(Corresponding author: Jingyan Wang.) Dinh Danh Le and Jingyan Wang are with the School of Electronic Engineering, Faculty of Engineering and Computing, Dublin City University, D09 V209 Dublin, Ireland, and also with the Hong Duc University (HDU), Thanh Hoa 40000-42000, Vietnam (e-mail: dinhdanh.le@dcu.ie; jingyan.wang@dcu.ie).

Funding Information:
The authors would like to acknowledge the support of the Science Foundation of Ireland (SFI) under the US-Ireland R&D Partnership Programme (SFI project code 15/US C2C/I3132), the SFI Centres CONNECT and IPIC.

Publisher Copyright:
© 1983-2012 IEEE.

Copyright:
Copyright 2020 Elsevier B.V., All rights reserved.

Keywords

Agile reconfigurability
data center networks
optical networking
traffic scheduling

Access to Document

10.1109/JLT.2020.2999778

09107450Final published version, 2.5 MBLicence: CC BY

Cite this

@article{c3b9281e4ce94e06b7c44f688eb06e81,

title = "AgileDCN: An Agile Reconfigurable Optical Data Center Network Architecture",

abstract = "This paper presents a detailed examination of a novel data center network (DCN) that can satisfy the high capacity and low latency requirements of modern cloud computing applications. This reconfigurable architecture called AgileDCN uses fast-switching optical components with a centralized control function and workload scheduler. By providing a highly flexible optical network fabric between server racks, very high network efficiencies can be achieved even under imbalanced loading patterns. Our simulation results show that, at high (70%) loads, TCP flow completion times in the AgileDCN are significantly lower than in an equivalent electronic leaf-spine network.",

keywords = "Agile reconfigurability, data center networks, optical networking, traffic scheduling",

author = "Le, {Dinh Danh} and Barry, {Liam P.} and Kilper, {Daniel C.} and Philip Perry and Jingyan Wang and Conor McArdle",

note = "Funding Information: This work was supported in part by the National Science Foundation (NSF) under Grant CNS-1827923, Grant CNS-1650669, Grant CNS-1737453, and Grant EEC-0812072, and in part by the Science Foundation of Ireland (SFI) under the US-Ireland R&D Partnership Programme (SFI project code 15/USC2C/I3132), the SFI Centres CONNECT, and IPIC. (Corresponding author: Jingyan Wang.) Dinh Danh Le and Jingyan Wang are with the School of Electronic Engineering, Faculty of Engineering and Computing, Dublin City University, D09 V209 Dublin, Ireland, and also with the Hong Duc University (HDU), Thanh Hoa 40000-42000, Vietnam (e-mail: dinhdanh.le@dcu.ie; jingyan.wang@dcu.ie). Funding Information: The authors would like to acknowledge the support of the Science Foundation of Ireland (SFI) under the US-Ireland R&D Partnership Programme (SFI project code 15/US C2C/I3132), the SFI Centres CONNECT and IPIC. Publisher Copyright: {\textcopyright} 1983-2012 IEEE. Copyright: Copyright 2020 Elsevier B.V., All rights reserved.",

year = "2020",

month = sep,

day = "15",

doi = "10.1109/JLT.2020.2999778",

language = "English",

volume = "38",

pages = "4922--4934",

journal = "Journal of Lightwave Technology",

issn = "0733-8724",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "18",

}

TY - JOUR

T1 - AgileDCN

T2 - An Agile Reconfigurable Optical Data Center Network Architecture

AU - Le, Dinh Danh

AU - Barry, Liam P.

AU - Kilper, Daniel C.

AU - Perry, Philip

AU - Wang, Jingyan

AU - McArdle, Conor

N1 - Funding Information: This work was supported in part by the National Science Foundation (NSF) under Grant CNS-1827923, Grant CNS-1650669, Grant CNS-1737453, and Grant EEC-0812072, and in part by the Science Foundation of Ireland (SFI) under the US-Ireland R&D Partnership Programme (SFI project code 15/USC2C/I3132), the SFI Centres CONNECT, and IPIC. (Corresponding author: Jingyan Wang.) Dinh Danh Le and Jingyan Wang are with the School of Electronic Engineering, Faculty of Engineering and Computing, Dublin City University, D09 V209 Dublin, Ireland, and also with the Hong Duc University (HDU), Thanh Hoa 40000-42000, Vietnam (e-mail: dinhdanh.le@dcu.ie; jingyan.wang@dcu.ie). Funding Information: The authors would like to acknowledge the support of the Science Foundation of Ireland (SFI) under the US-Ireland R&D Partnership Programme (SFI project code 15/US C2C/I3132), the SFI Centres CONNECT and IPIC. Publisher Copyright: © 1983-2012 IEEE. Copyright: Copyright 2020 Elsevier B.V., All rights reserved.

PY - 2020/9/15

Y1 - 2020/9/15

N2 - This paper presents a detailed examination of a novel data center network (DCN) that can satisfy the high capacity and low latency requirements of modern cloud computing applications. This reconfigurable architecture called AgileDCN uses fast-switching optical components with a centralized control function and workload scheduler. By providing a highly flexible optical network fabric between server racks, very high network efficiencies can be achieved even under imbalanced loading patterns. Our simulation results show that, at high (70%) loads, TCP flow completion times in the AgileDCN are significantly lower than in an equivalent electronic leaf-spine network.

AB - This paper presents a detailed examination of a novel data center network (DCN) that can satisfy the high capacity and low latency requirements of modern cloud computing applications. This reconfigurable architecture called AgileDCN uses fast-switching optical components with a centralized control function and workload scheduler. By providing a highly flexible optical network fabric between server racks, very high network efficiencies can be achieved even under imbalanced loading patterns. Our simulation results show that, at high (70%) loads, TCP flow completion times in the AgileDCN are significantly lower than in an equivalent electronic leaf-spine network.

KW - Agile reconfigurability

KW - data center networks

KW - optical networking

KW - traffic scheduling

UR - http://www.scopus.com/inward/record.url?scp=85090928305&partnerID=8YFLogxK

U2 - 10.1109/JLT.2020.2999778

DO - 10.1109/JLT.2020.2999778

M3 - Article

AN - SCOPUS:85090928305

VL - 38

SP - 4922

EP - 4934

JO - Journal of Lightwave Technology

JF - Journal of Lightwave Technology

SN - 0733-8724

IS - 18

M1 - 9107450

ER -

AgileDCN: An Agile Reconfigurable Optical Data Center Network Architecture

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this