Container Density Improvements with Dynamic Memory Extension using NAND Flash

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Authors

External Research Organisations

  • Diablo Technologies Inc.
  • IBM Research
  • Delft University of Technology
View graph of relations

Details

Original languageEnglish
Title of host publicationApSys 2018, Proceedings of the 9th Asia-Pacific Workshop on Systems
PublisherAssociation for Computing Machinery (ACM)
Pages1-8
Number of pages8
ISBN (electronic)9781450360067
ISBN (print)9781450360067
Publication statusPublished - 2018
Externally publishedYes
Event9th ACM SIGOPS Asia-Pacific Workshop on Systems - , Korea, Republic of
Duration: 27 Aug 201828 Aug 2018

Publication series

NameProceedings of the 9th Asia-Pacific Workshop on Systems, APSys 2018

Abstract

While containers efficiently implement the idea of operating-system-level application virtualization, they are often insufficient to increase the server utilization to a desirable level. The reason is that in practice many containerized applications experience a limited amount of load while there are few containers with a high load. In such a scenario, the virtual memory management system can become the limiting factor to container density even though the working set of active containers would fit into main memory. In this paper, we describe and evaluate a system for transparently moving memory pages in and out of DRAM and to a NAND Flash medium which is attached through the memory bus. This technique, called Diablo Memory Expansion (DMX), operates on a prediction model and is able to relieve the pressure on the memory system. We present a benchmark for container density and show that even under an overall constant workload, adding additional containers adversely affects performance-critical applications in Docker. When using the DMX technology of the Memory1 system, however, the performance of the critical workload remains stable.

ASJC Scopus subject areas

Cite this

Container Density Improvements with Dynamic Memory Extension using NAND Flash. / Rellermeyer, Jan; Amer, Maher; Smutzer, Richard et al.
ApSys 2018, Proceedings of the 9th Asia-Pacific Workshop on Systems. Association for Computing Machinery (ACM), 2018. p. 1-8 10 (Proceedings of the 9th Asia-Pacific Workshop on Systems, APSys 2018).

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Rellermeyer, J, Amer, M, Smutzer, R & Rajamani, K 2018, Container Density Improvements with Dynamic Memory Extension using NAND Flash. in ApSys 2018, Proceedings of the 9th Asia-Pacific Workshop on Systems., 10, Proceedings of the 9th Asia-Pacific Workshop on Systems, APSys 2018, Association for Computing Machinery (ACM), pp. 1-8, 9th ACM SIGOPS Asia-Pacific Workshop on Systems, Korea, Republic of, 27 Aug 2018. https://doi.org/10.1145/3265723.3265740
Rellermeyer, J., Amer, M., Smutzer, R., & Rajamani, K. (2018). Container Density Improvements with Dynamic Memory Extension using NAND Flash. In ApSys 2018, Proceedings of the 9th Asia-Pacific Workshop on Systems (pp. 1-8). Article 10 (Proceedings of the 9th Asia-Pacific Workshop on Systems, APSys 2018). Association for Computing Machinery (ACM). https://doi.org/10.1145/3265723.3265740
Rellermeyer J, Amer M, Smutzer R, Rajamani K. Container Density Improvements with Dynamic Memory Extension using NAND Flash. In ApSys 2018, Proceedings of the 9th Asia-Pacific Workshop on Systems. Association for Computing Machinery (ACM). 2018. p. 1-8. 10. (Proceedings of the 9th Asia-Pacific Workshop on Systems, APSys 2018). doi: 10.1145/3265723.3265740
Rellermeyer, Jan ; Amer, Maher ; Smutzer, Richard et al. / Container Density Improvements with Dynamic Memory Extension using NAND Flash. ApSys 2018, Proceedings of the 9th Asia-Pacific Workshop on Systems. Association for Computing Machinery (ACM), 2018. pp. 1-8 (Proceedings of the 9th Asia-Pacific Workshop on Systems, APSys 2018).
Download
@inproceedings{ca000ae523634d04910c7b032059b1b9,
title = "Container Density Improvements with Dynamic Memory Extension using NAND Flash",
abstract = "While containers efficiently implement the idea of operating-system-level application virtualization, they are often insufficient to increase the server utilization to a desirable level. The reason is that in practice many containerized applications experience a limited amount of load while there are few containers with a high load. In such a scenario, the virtual memory management system can become the limiting factor to container density even though the working set of active containers would fit into main memory. In this paper, we describe and evaluate a system for transparently moving memory pages in and out of DRAM and to a NAND Flash medium which is attached through the memory bus. This technique, called Diablo Memory Expansion (DMX), operates on a prediction model and is able to relieve the pressure on the memory system. We present a benchmark for container density and show that even under an overall constant workload, adding additional containers adversely affects performance-critical applications in Docker. When using the DMX technology of the Memory1 system, however, the performance of the critical workload remains stable.",
author = "Jan Rellermeyer and Maher Amer and Richard Smutzer and Karthick Rajamani",
note = "Publisher Copyright: {\textcopyright} 2018 Copyright held by the owner/author(s).; 9th ACM SIGOPS Asia-Pacific Workshop on Systems ; Conference date: 27-08-2018 Through 28-08-2018",
year = "2018",
doi = "10.1145/3265723.3265740",
language = "English",
isbn = "9781450360067",
series = "Proceedings of the 9th Asia-Pacific Workshop on Systems, APSys 2018",
publisher = "Association for Computing Machinery (ACM)",
pages = "1--8",
booktitle = "ApSys 2018, Proceedings of the 9th Asia-Pacific Workshop on Systems",
address = "United States",

}

Download

TY - GEN

T1 - Container Density Improvements with Dynamic Memory Extension using NAND Flash

AU - Rellermeyer, Jan

AU - Amer, Maher

AU - Smutzer, Richard

AU - Rajamani, Karthick

N1 - Publisher Copyright: © 2018 Copyright held by the owner/author(s).

PY - 2018

Y1 - 2018

N2 - While containers efficiently implement the idea of operating-system-level application virtualization, they are often insufficient to increase the server utilization to a desirable level. The reason is that in practice many containerized applications experience a limited amount of load while there are few containers with a high load. In such a scenario, the virtual memory management system can become the limiting factor to container density even though the working set of active containers would fit into main memory. In this paper, we describe and evaluate a system for transparently moving memory pages in and out of DRAM and to a NAND Flash medium which is attached through the memory bus. This technique, called Diablo Memory Expansion (DMX), operates on a prediction model and is able to relieve the pressure on the memory system. We present a benchmark for container density and show that even under an overall constant workload, adding additional containers adversely affects performance-critical applications in Docker. When using the DMX technology of the Memory1 system, however, the performance of the critical workload remains stable.

AB - While containers efficiently implement the idea of operating-system-level application virtualization, they are often insufficient to increase the server utilization to a desirable level. The reason is that in practice many containerized applications experience a limited amount of load while there are few containers with a high load. In such a scenario, the virtual memory management system can become the limiting factor to container density even though the working set of active containers would fit into main memory. In this paper, we describe and evaluate a system for transparently moving memory pages in and out of DRAM and to a NAND Flash medium which is attached through the memory bus. This technique, called Diablo Memory Expansion (DMX), operates on a prediction model and is able to relieve the pressure on the memory system. We present a benchmark for container density and show that even under an overall constant workload, adding additional containers adversely affects performance-critical applications in Docker. When using the DMX technology of the Memory1 system, however, the performance of the critical workload remains stable.

UR - http://www.scopus.com/inward/record.url?scp=85055894235&partnerID=8YFLogxK

U2 - 10.1145/3265723.3265740

DO - 10.1145/3265723.3265740

M3 - Conference contribution

SN - 9781450360067

T3 - Proceedings of the 9th Asia-Pacific Workshop on Systems, APSys 2018

SP - 1

EP - 8

BT - ApSys 2018, Proceedings of the 9th Asia-Pacific Workshop on Systems

PB - Association for Computing Machinery (ACM)

T2 - 9th ACM SIGOPS Asia-Pacific Workshop on Systems

Y2 - 27 August 2018 through 28 August 2018

ER -

By the same author(s)