PipeDevice: A Hardware-Software Co-Design Approach to Intra-Host Container Communication

Qiang Su; Chuanwen Wang; Zhixiong Niu; Ran Shu; Peng Cheng; Yongqiang Xiong; Dongsu Han; Chun Jason Xue; Hong Xu

doi:10.1145/3546037.3546054

PipeDevice: A Hardware-Software Co-Design Approach to Intra-Host Container Communication

Qiang Su, Chuanwen Wang, Zhixiong Niu, Ran Shu, Peng Cheng, Yongqiang Xiong, Dongsu Han, Chun Jason Xue, Hong Xu

Department of Computer Science

Research output: Conference Papers › Poster › peer-review

1 Citation (Scopus)

Abstract

Containers have become prevalent in public clouds due to the performance, portability, and deployment benefits compared to virtual machines [1, 8, 14]. They support a wide variety of workloads, from microservices to data analytics and machine learning. Containerized applications often entail extensive bulky data transfers to exchange intermediate results of data processing among peers. Examples include the shuffle stage in MapReduce jobs [3, 15, 16, 23] and the model update process with parameter server and all reduce in distributed machine learning [5, 19, 32, 35].

Original language	English
Pages	28-30
Number of pages	3
DOIs	https://doi.org/10.1145/3546037.3546054
Publication status	Published - 2022
Event	ACM SIGCOMM 2022 - The Beurs van Berlage conference venue, Amsterdam, Netherlands Duration: 22 Aug 2022 → 26 Aug 2022 https://conferences.sigcomm.org/sigcomm/2022/

Conference

Conference	ACM SIGCOMM 2022
Abbreviated title	SIGCOMM 2022
Place	Netherlands
City	Amsterdam
Period	22/08/22 → 26/08/22
Internet address	https://conferences.sigcomm.org/sigcomm/2022/

Bibliographical note

Research Unit(s) information for this publication is provided by the author(s) concerned.

Research Keywords

Container Communication
Hardware-Software Co-Design

Access Output

10.1145/3546037.3546054

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@conference{27bceb274bb54bb4ba00d7e0e39dcf8d,

title = "PipeDevice: A Hardware-Software Co-Design Approach to Intra-Host Container Communication",

abstract = "Containers have become prevalent in public clouds due to the performance, portability, and deployment benefits compared to virtual machines [1, 8, 14]. They support a wide variety of workloads, from microservices to data analytics and machine learning. Containerized applications often entail extensive bulky data transfers to exchange intermediate results of data processing among peers. Examples include the shuffle stage in MapReduce jobs [3, 15, 16, 23] and the model update process with parameter server and all reduce in distributed machine learning [5, 19, 32, 35].",

keywords = "Container Communication, Hardware-Software Co-Design",

author = "Qiang Su and Chuanwen Wang and Zhixiong Niu and Ran Shu and Peng Cheng and Yongqiang Xiong and Dongsu Han and Xue, {Chun Jason} and Hong Xu",

note = "Research Unit(s) information for this publication is provided by the author(s) concerned.; ACM SIGCOMM 2022, SIGCOMM 2022 ; Conference date: 22-08-2022 Through 26-08-2022",

year = "2022",

doi = "10.1145/3546037.3546054",

language = "English",

pages = "28--30",

url = "https://conferences.sigcomm.org/sigcomm/2022/",

}

TY - CONF

T1 - PipeDevice

T2 - ACM SIGCOMM 2022

AU - Su, Qiang

AU - Wang, Chuanwen

AU - Niu, Zhixiong

AU - Shu, Ran

AU - Cheng, Peng

AU - Xiong, Yongqiang

AU - Han, Dongsu

AU - Xue, Chun Jason

AU - Xu, Hong

N1 - Research Unit(s) information for this publication is provided by the author(s) concerned.

PY - 2022

Y1 - 2022

N2 - Containers have become prevalent in public clouds due to the performance, portability, and deployment benefits compared to virtual machines [1, 8, 14]. They support a wide variety of workloads, from microservices to data analytics and machine learning. Containerized applications often entail extensive bulky data transfers to exchange intermediate results of data processing among peers. Examples include the shuffle stage in MapReduce jobs [3, 15, 16, 23] and the model update process with parameter server and all reduce in distributed machine learning [5, 19, 32, 35].

AB - Containers have become prevalent in public clouds due to the performance, portability, and deployment benefits compared to virtual machines [1, 8, 14]. They support a wide variety of workloads, from microservices to data analytics and machine learning. Containerized applications often entail extensive bulky data transfers to exchange intermediate results of data processing among peers. Examples include the shuffle stage in MapReduce jobs [3, 15, 16, 23] and the model update process with parameter server and all reduce in distributed machine learning [5, 19, 32, 35].

KW - Container Communication

KW - Hardware-Software Co-Design

UR - http://www.scopus.com/inward/record.url?scp=85141698632&partnerID=8YFLogxK

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85141698632&origin=recordpage

U2 - 10.1145/3546037.3546054

DO - 10.1145/3546037.3546054

M3 - Poster

SP - 28

EP - 30

Y2 - 22 August 2022 through 26 August 2022

ER -

PipeDevice: A Hardware-Software Co-Design Approach to Intra-Host Container Communication

Abstract

Conference

Bibliographical note

Research Keywords

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this