Fast Function Instantiation with Alternate Virtualization Approaches

Vivek Jain; Shixiong Qi; K. K. Ramakrishnan

doi:10.1109/LANMAN52105.2021.9478808

Fast Function Instantiation with Alternate Virtualization Approaches

Vivek Jain, Shixiong Qi, K. K. Ramakrishnan

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Scopus citations

Abstract

This paper focuses on the need for emerging domains such as serverless and in-network computing, where applications are often hosted on virtualized compute instances (e.g., containers and unikernels), to have applications startup as quickly as possible. We provide a qualitative and quantitative analysis of containers and unikernels with regard to the startup time. We analyze these in-depth and identify the key components and their impact under scale on the startup latency. We study how startup time scales as we launch multiple instances concurrently. We study the contribution of popular Container Networking Interfaces (CNIs), to the startup time.

Original language	English
Title of host publication	27th IEEE International Symposium on Local and Metropolitan Area Networks, LANMAN 2021
ISBN (Electronic)	9781665445795
DOIs	https://doi.org/10.1109/LANMAN52105.2021.9478808
State	Published - Jul 12 2021
Event	27th IEEE International Symposium on Local and Metropolitan Area Networks, LANMAN 2021 - Virtual, Online Duration: Jul 12 2021 → Jul 13 2021

Publication series

Name	IEEE Workshop on Local and Metropolitan Area Networks
Volume	2021-July
ISSN (Print)	1944-0367
ISSN (Electronic)	1944-0375

Conference

Conference	27th IEEE International Symposium on Local and Metropolitan Area Networks, LANMAN 2021
City	Virtual, Online
Period	7/12/21 → 7/13/21

Bibliographical note

Publisher Copyright:
© 2021 IEEE.

Funding

We thank the US NSF for their generous support of this work through grant CNS-1763929. This research was also sponsored by the OUSD(R&E)/RT&L and was accomplished under Cooperative Agreement Number W911NF-20-2-0267. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the ARL and OUSD(R&E)/RT&L or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation herein. We also thank the anonymous reviewers for their valuable feedback. ACKNOWLEDGMENT We thank the US NSF for their generous support of this work through grant CNS-1763929. This research was also sponsored by the OUSD(R&E)/RT&L and was accomplished under Cooperative Agreement Number W911NF-20-2-0267. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the ARL and OUSD(R&E)/RT&L or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation herein. We also thank the anonymous reviewers for their valuable feedback.

Funders	Funder number
OUSD
National Science Foundation Arctic Social Science Program	W911NF-20-2-0267, CNS-1763929
Army Research Laboratory

Keywords

cold start
Container
NFV
startup
Unikernel

ASJC Scopus subject areas

Computer Networks and Communications
Software
Electrical and Electronic Engineering
Communication

Access to Document

10.1109/LANMAN52105.2021.9478808

Cite this

@inproceedings{6ec3a53439f74403b676c07f3579b735,

title = "Fast Function Instantiation with Alternate Virtualization Approaches",

abstract = "This paper focuses on the need for emerging domains such as serverless and in-network computing, where applications are often hosted on virtualized compute instances (e.g., containers and unikernels), to have applications startup as quickly as possible. We provide a qualitative and quantitative analysis of containers and unikernels with regard to the startup time. We analyze these in-depth and identify the key components and their impact under scale on the startup latency. We study how startup time scales as we launch multiple instances concurrently. We study the contribution of popular Container Networking Interfaces (CNIs), to the startup time.",

keywords = "cold start, Container, NFV, startup, Unikernel",

author = "Vivek Jain and Shixiong Qi and Ramakrishnan, \{K. K.\}",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 27th IEEE International Symposium on Local and Metropolitan Area Networks, LANMAN 2021 ; Conference date: 12-07-2021 Through 13-07-2021",

year = "2021",

month = jul,

day = "12",

doi = "10.1109/LANMAN52105.2021.9478808",

language = "English",

series = "IEEE Workshop on Local and Metropolitan Area Networks",

booktitle = "27th IEEE International Symposium on Local and Metropolitan Area Networks, LANMAN 2021",

}

Jain, V, Qi, S & Ramakrishnan, KK 2021, Fast Function Instantiation with Alternate Virtualization Approaches. in 27th IEEE International Symposium on Local and Metropolitan Area Networks, LANMAN 2021., 9478808, IEEE Workshop on Local and Metropolitan Area Networks, vol. 2021-July, 27th IEEE International Symposium on Local and Metropolitan Area Networks, LANMAN 2021, Virtual, Online, 7/12/21. https://doi.org/10.1109/LANMAN52105.2021.9478808

Fast Function Instantiation with Alternate Virtualization Approaches. / Jain, Vivek; Qi, Shixiong; Ramakrishnan, K. K.
27th IEEE International Symposium on Local and Metropolitan Area Networks, LANMAN 2021. 2021. 9478808 (IEEE Workshop on Local and Metropolitan Area Networks; Vol. 2021-July).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Fast Function Instantiation with Alternate Virtualization Approaches

AU - Jain, Vivek

AU - Qi, Shixiong

AU - Ramakrishnan, K. K.

PY - 2021/7/12

Y1 - 2021/7/12

N2 - This paper focuses on the need for emerging domains such as serverless and in-network computing, where applications are often hosted on virtualized compute instances (e.g., containers and unikernels), to have applications startup as quickly as possible. We provide a qualitative and quantitative analysis of containers and unikernels with regard to the startup time. We analyze these in-depth and identify the key components and their impact under scale on the startup latency. We study how startup time scales as we launch multiple instances concurrently. We study the contribution of popular Container Networking Interfaces (CNIs), to the startup time.

AB - This paper focuses on the need for emerging domains such as serverless and in-network computing, where applications are often hosted on virtualized compute instances (e.g., containers and unikernels), to have applications startup as quickly as possible. We provide a qualitative and quantitative analysis of containers and unikernels with regard to the startup time. We analyze these in-depth and identify the key components and their impact under scale on the startup latency. We study how startup time scales as we launch multiple instances concurrently. We study the contribution of popular Container Networking Interfaces (CNIs), to the startup time.

KW - cold start

KW - Container

KW - NFV

KW - startup

KW - Unikernel

UR - http://www.scopus.com/inward/record.url?scp=85113872645&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85113872645&partnerID=8YFLogxK

U2 - 10.1109/LANMAN52105.2021.9478808

DO - 10.1109/LANMAN52105.2021.9478808

M3 - Conference contribution

AN - SCOPUS:85113872645

T3 - IEEE Workshop on Local and Metropolitan Area Networks

BT - 27th IEEE International Symposium on Local and Metropolitan Area Networks, LANMAN 2021

T2 - 27th IEEE International Symposium on Local and Metropolitan Area Networks, LANMAN 2021

Y2 - 12 July 2021 through 13 July 2021

ER -

Fast Function Instantiation with Alternate Virtualization Approaches

Abstract

Publication series

Conference

Bibliographical note

Funding

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this