S-RAC

SSD friendly caching for data center workloads

Yuanjiang Ni, Ji Jiang, Dejun Jiang, Xiaosong Ma, Jin Xiong, Yuangang Wang

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

5 Citations (Scopus)

Abstract

Current data-center applications tend to process increasingly large volume of data sets. The caching effect of page cache is reduced by its limited capacity. Emerging flashbased solid state drives (SSD) have latency and price advantages compared to hard disk and DRAM. Thus, SSD-based caching is widely deployed in data centers. However, SSD caching faces two challenges. First, SSD has limited write endurance, which requires cache manager to reduce write amount to SSD. Second, data-center workloads exhibit a diverse I/O access patterns, which requires one to figure out SSD caching friendly access patterns. This paper first classifies 6 I/O access patterns among 32 data-center workloads using a cost-benefit analysis. We derive implications for designing SSD cache from analyzing the access patterns. We then propose an SSD cache manager S-RAC with re-Adding blocks and ghost cache adaptation to retain SSD friendly blocks in SSD. The experimental evaluation shows the efficiency of S-RAC in reducing SSD write amount while improving/ maintaining cache hit ratio.

Original languageEnglish
Title of host publicationSYSTOR 2016 - Proceedings of the 9th ACM International Systems and Storage Conference
PublisherAssociation for Computing Machinery, Inc
ISBN (Electronic)9781450343817
DOIs
Publication statusPublished - 6 Jun 2016
Event9th ACM International Systems and Storage Conference, SYSTOR 2016 - Haifa, Israel
Duration: 6 Jun 20168 Jun 2016

Other

Other9th ACM International Systems and Storage Conference, SYSTOR 2016
CountryIsrael
CityHaifa
Period6/6/168/6/16

Fingerprint

Managers
Cost benefit analysis
Dynamic random access storage
Hard disk storage
Durability

Keywords

  • Cache hit ratio
  • SSD Caching
  • Write endurance

ASJC Scopus subject areas

  • Computer Science Applications
  • Electrical and Electronic Engineering
  • Hardware and Architecture
  • Software

Cite this

Ni, Y., Jiang, J., Jiang, D., Ma, X., Xiong, J., & Wang, Y. (2016). S-RAC: SSD friendly caching for data center workloads. In SYSTOR 2016 - Proceedings of the 9th ACM International Systems and Storage Conference [2928284] Association for Computing Machinery, Inc. https://doi.org/10.1145/2928275.2928284

S-RAC : SSD friendly caching for data center workloads. / Ni, Yuanjiang; Jiang, Ji; Jiang, Dejun; Ma, Xiaosong; Xiong, Jin; Wang, Yuangang.

SYSTOR 2016 - Proceedings of the 9th ACM International Systems and Storage Conference. Association for Computing Machinery, Inc, 2016. 2928284.

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

Ni, Y, Jiang, J, Jiang, D, Ma, X, Xiong, J & Wang, Y 2016, S-RAC: SSD friendly caching for data center workloads. in SYSTOR 2016 - Proceedings of the 9th ACM International Systems and Storage Conference., 2928284, Association for Computing Machinery, Inc, 9th ACM International Systems and Storage Conference, SYSTOR 2016, Haifa, Israel, 6/6/16. https://doi.org/10.1145/2928275.2928284
Ni Y, Jiang J, Jiang D, Ma X, Xiong J, Wang Y. S-RAC: SSD friendly caching for data center workloads. In SYSTOR 2016 - Proceedings of the 9th ACM International Systems and Storage Conference. Association for Computing Machinery, Inc. 2016. 2928284 https://doi.org/10.1145/2928275.2928284
Ni, Yuanjiang ; Jiang, Ji ; Jiang, Dejun ; Ma, Xiaosong ; Xiong, Jin ; Wang, Yuangang. / S-RAC : SSD friendly caching for data center workloads. SYSTOR 2016 - Proceedings of the 9th ACM International Systems and Storage Conference. Association for Computing Machinery, Inc, 2016.
@inproceedings{b3d05cb6b79f4111969c49a5a496ba0a,
title = "S-RAC: SSD friendly caching for data center workloads",
abstract = "Current data-center applications tend to process increasingly large volume of data sets. The caching effect of page cache is reduced by its limited capacity. Emerging flashbased solid state drives (SSD) have latency and price advantages compared to hard disk and DRAM. Thus, SSD-based caching is widely deployed in data centers. However, SSD caching faces two challenges. First, SSD has limited write endurance, which requires cache manager to reduce write amount to SSD. Second, data-center workloads exhibit a diverse I/O access patterns, which requires one to figure out SSD caching friendly access patterns. This paper first classifies 6 I/O access patterns among 32 data-center workloads using a cost-benefit analysis. We derive implications for designing SSD cache from analyzing the access patterns. We then propose an SSD cache manager S-RAC with re-Adding blocks and ghost cache adaptation to retain SSD friendly blocks in SSD. The experimental evaluation shows the efficiency of S-RAC in reducing SSD write amount while improving/ maintaining cache hit ratio.",
keywords = "Cache hit ratio, SSD Caching, Write endurance",
author = "Yuanjiang Ni and Ji Jiang and Dejun Jiang and Xiaosong Ma and Jin Xiong and Yuangang Wang",
year = "2016",
month = "6",
day = "6",
doi = "10.1145/2928275.2928284",
language = "English",
booktitle = "SYSTOR 2016 - Proceedings of the 9th ACM International Systems and Storage Conference",
publisher = "Association for Computing Machinery, Inc",

}

TY - GEN

T1 - S-RAC

T2 - SSD friendly caching for data center workloads

AU - Ni, Yuanjiang

AU - Jiang, Ji

AU - Jiang, Dejun

AU - Ma, Xiaosong

AU - Xiong, Jin

AU - Wang, Yuangang

PY - 2016/6/6

Y1 - 2016/6/6

N2 - Current data-center applications tend to process increasingly large volume of data sets. The caching effect of page cache is reduced by its limited capacity. Emerging flashbased solid state drives (SSD) have latency and price advantages compared to hard disk and DRAM. Thus, SSD-based caching is widely deployed in data centers. However, SSD caching faces two challenges. First, SSD has limited write endurance, which requires cache manager to reduce write amount to SSD. Second, data-center workloads exhibit a diverse I/O access patterns, which requires one to figure out SSD caching friendly access patterns. This paper first classifies 6 I/O access patterns among 32 data-center workloads using a cost-benefit analysis. We derive implications for designing SSD cache from analyzing the access patterns. We then propose an SSD cache manager S-RAC with re-Adding blocks and ghost cache adaptation to retain SSD friendly blocks in SSD. The experimental evaluation shows the efficiency of S-RAC in reducing SSD write amount while improving/ maintaining cache hit ratio.

AB - Current data-center applications tend to process increasingly large volume of data sets. The caching effect of page cache is reduced by its limited capacity. Emerging flashbased solid state drives (SSD) have latency and price advantages compared to hard disk and DRAM. Thus, SSD-based caching is widely deployed in data centers. However, SSD caching faces two challenges. First, SSD has limited write endurance, which requires cache manager to reduce write amount to SSD. Second, data-center workloads exhibit a diverse I/O access patterns, which requires one to figure out SSD caching friendly access patterns. This paper first classifies 6 I/O access patterns among 32 data-center workloads using a cost-benefit analysis. We derive implications for designing SSD cache from analyzing the access patterns. We then propose an SSD cache manager S-RAC with re-Adding blocks and ghost cache adaptation to retain SSD friendly blocks in SSD. The experimental evaluation shows the efficiency of S-RAC in reducing SSD write amount while improving/ maintaining cache hit ratio.

KW - Cache hit ratio

KW - SSD Caching

KW - Write endurance

UR - http://www.scopus.com/inward/record.url?scp=84978906720&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84978906720&partnerID=8YFLogxK

U2 - 10.1145/2928275.2928284

DO - 10.1145/2928275.2928284

M3 - Conference contribution

BT - SYSTOR 2016 - Proceedings of the 9th ACM International Systems and Storage Conference

PB - Association for Computing Machinery, Inc

ER -