Efficient multi-version storage engine for main memory data store

Jinwei Guo; Bing Xiao; Peng Cai; Weining Qian; Aoying Zhou

doi:10.1007/978-3-319-68786-5_17

Efficient multi-version storage engine for main memory data store

Jinwei Guo
, Bing Xiao
, Peng Cai^*
, Weining Qian
, Aoying Zhou

^*Corresponding author for this work

School of Data Science and Engineering

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

Abstract

Multi-version storage engine is the fundamental component of modern main memory data store using the popular multiple version concurrency control (MVCC). The straightforward implementation of storage engine is to use a linked list to store multiple versions of an object. A read operation has to traverse the list for the specified version, which incurs pointer chasing. An optimization method implemented in HyPer is to store the current version in the object header, which is friendly to read the latest snapshot of data. However, a read operation still needs one extra pointer chasing in memory when accessing an object being updated. In this paper, we propose an efficient multi-version storage (EMS), a new storage engine for main memory data store. EMS embeds two latest versions in each object header, so that it can avoid the overhead of traversal of version list, especially in the update-intensive scenario. We present an implementation mechanism of widely used snapshot isolation level over EMS. The experimental results demonstrate that EMS outperforms the exiting multi-version storage engine of well-known main memory data stores in terms of throughput without excessive memory consumption.

Original language	English
Title of host publication	Web Information Systems Engineering – WISE 2017 - 18th International Conference, Proceedings
Editors	Weijia Jia, Andrey Klimenko, Fedor Dzerzhinskiy, Stanislav V. Klimenko, Lu Chen, Qing Li, Yunjun Gao, Athman Bouguettaya, Xiangliang Zhang
Publisher	Springer Verlag
Pages	205-220
Number of pages	16
ISBN (Print)	9783319687858
DOIs	https://doi.org/10.1007/978-3-319-68786-5_17
State	Published - 2017
Event	18th International Conference on Web Information Systems Engineering, WISE 2017 - Puschino, Russian Federation Duration: 7 Oct 2017 → 11 Oct 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10570 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	18th International Conference on Web Information Systems Engineering, WISE 2017
Country/Territory	Russian Federation
City	Puschino
Period	7/10/17 → 11/10/17

Keywords

Main memory
Multi-version storage
Snapshot isolation

Access to Document

10.1007/978-3-319-68786-5_17

Cite this

Guo, J., Xiao, B., Cai, P., Qian, W., & Zhou, A. (2017). Efficient multi-version storage engine for main memory data store. In W. Jia, A. Klimenko, F. Dzerzhinskiy, S. V. Klimenko, L. Chen, Q. Li, Y. Gao, A. Bouguettaya, & X. Zhang (Eds.), Web Information Systems Engineering – WISE 2017 - 18th International Conference, Proceedings (pp. 205-220). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10570 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-68786-5_17

Guo, Jinwei ; Xiao, Bing ; Cai, Peng et al. / Efficient multi-version storage engine for main memory data store. Web Information Systems Engineering – WISE 2017 - 18th International Conference, Proceedings. editor / Weijia Jia ; Andrey Klimenko ; Fedor Dzerzhinskiy ; Stanislav V. Klimenko ; Lu Chen ; Qing Li ; Yunjun Gao ; Athman Bouguettaya ; Xiangliang Zhang. Springer Verlag, 2017. pp. 205-220 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{542e6561abcd4801b53dbf4bc72877fa,

title = "Efficient multi-version storage engine for main memory data store",

abstract = "Multi-version storage engine is the fundamental component of modern main memory data store using the popular multiple version concurrency control (MVCC). The straightforward implementation of storage engine is to use a linked list to store multiple versions of an object. A read operation has to traverse the list for the specified version, which incurs pointer chasing. An optimization method implemented in HyPer is to store the current version in the object header, which is friendly to read the latest snapshot of data. However, a read operation still needs one extra pointer chasing in memory when accessing an object being updated. In this paper, we propose an efficient multi-version storage (EMS), a new storage engine for main memory data store. EMS embeds two latest versions in each object header, so that it can avoid the overhead of traversal of version list, especially in the update-intensive scenario. We present an implementation mechanism of widely used snapshot isolation level over EMS. The experimental results demonstrate that EMS outperforms the exiting multi-version storage engine of well-known main memory data stores in terms of throughput without excessive memory consumption.",

keywords = "Main memory, Multi-version storage, Snapshot isolation",

author = "Jinwei Guo and Bing Xiao and Peng Cai and Weining Qian and Aoying Zhou",

note = "Publisher Copyright: {\textcopyright} 2017, Springer International Publishing AG.; 18th International Conference on Web Information Systems Engineering, WISE 2017 ; Conference date: 07-10-2017 Through 11-10-2017",

year = "2017",

doi = "10.1007/978-3-319-68786-5\_17",

language = "英语",

isbn = "9783319687858",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "205--220",

editor = "Weijia Jia and Andrey Klimenko and Fedor Dzerzhinskiy and Klimenko, \{Stanislav V.\} and Lu Chen and Qing Li and Yunjun Gao and Athman Bouguettaya and Xiangliang Zhang",

booktitle = "Web Information Systems Engineering – WISE 2017 - 18th International Conference, Proceedings",

address = "德国",

}

Guo, J, Xiao, B, Cai, P , Qian, W & Zhou, A 2017, Efficient multi-version storage engine for main memory data store. in W Jia, A Klimenko, F Dzerzhinskiy, SV Klimenko, L Chen, Q Li, Y Gao, A Bouguettaya & X Zhang (eds), Web Information Systems Engineering – WISE 2017 - 18th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10570 LNCS, Springer Verlag, pp. 205-220, 18th International Conference on Web Information Systems Engineering, WISE 2017, Puschino, Russian Federation, 7/10/17. https://doi.org/10.1007/978-3-319-68786-5_17

Efficient multi-version storage engine for main memory data store. / Guo, Jinwei; Xiao, Bing; Cai, Peng et al.
Web Information Systems Engineering – WISE 2017 - 18th International Conference, Proceedings. ed. / Weijia Jia; Andrey Klimenko; Fedor Dzerzhinskiy; Stanislav V. Klimenko; Lu Chen; Qing Li; Yunjun Gao; Athman Bouguettaya; Xiangliang Zhang. Springer Verlag, 2017. p. 205-220 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10570 LNCS).

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

TY - GEN

T1 - Efficient multi-version storage engine for main memory data store

AU - Guo, Jinwei

AU - Xiao, Bing

AU - Cai, Peng

AU - Qian, Weining

AU - Zhou, Aoying

PY - 2017

Y1 - 2017

N2 - Multi-version storage engine is the fundamental component of modern main memory data store using the popular multiple version concurrency control (MVCC). The straightforward implementation of storage engine is to use a linked list to store multiple versions of an object. A read operation has to traverse the list for the specified version, which incurs pointer chasing. An optimization method implemented in HyPer is to store the current version in the object header, which is friendly to read the latest snapshot of data. However, a read operation still needs one extra pointer chasing in memory when accessing an object being updated. In this paper, we propose an efficient multi-version storage (EMS), a new storage engine for main memory data store. EMS embeds two latest versions in each object header, so that it can avoid the overhead of traversal of version list, especially in the update-intensive scenario. We present an implementation mechanism of widely used snapshot isolation level over EMS. The experimental results demonstrate that EMS outperforms the exiting multi-version storage engine of well-known main memory data stores in terms of throughput without excessive memory consumption.

AB - Multi-version storage engine is the fundamental component of modern main memory data store using the popular multiple version concurrency control (MVCC). The straightforward implementation of storage engine is to use a linked list to store multiple versions of an object. A read operation has to traverse the list for the specified version, which incurs pointer chasing. An optimization method implemented in HyPer is to store the current version in the object header, which is friendly to read the latest snapshot of data. However, a read operation still needs one extra pointer chasing in memory when accessing an object being updated. In this paper, we propose an efficient multi-version storage (EMS), a new storage engine for main memory data store. EMS embeds two latest versions in each object header, so that it can avoid the overhead of traversal of version list, especially in the update-intensive scenario. We present an implementation mechanism of widely used snapshot isolation level over EMS. The experimental results demonstrate that EMS outperforms the exiting multi-version storage engine of well-known main memory data stores in terms of throughput without excessive memory consumption.

KW - Main memory

KW - Multi-version storage

KW - Snapshot isolation

UR - https://www.scopus.com/pages/publications/85031419810

U2 - 10.1007/978-3-319-68786-5_17

DO - 10.1007/978-3-319-68786-5_17

M3 - 会议稿件

AN - SCOPUS:85031419810

SN - 9783319687858

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 205

EP - 220

BT - Web Information Systems Engineering – WISE 2017 - 18th International Conference, Proceedings

A2 - Jia, Weijia

A2 - Klimenko, Andrey

A2 - Dzerzhinskiy, Fedor

A2 - Klimenko, Stanislav V.

A2 - Chen, Lu

A2 - Li, Qing

A2 - Gao, Yunjun

A2 - Bouguettaya, Athman

A2 - Zhang, Xiangliang

PB - Springer Verlag

T2 - 18th International Conference on Web Information Systems Engineering, WISE 2017

Y2 - 7 October 2017 through 11 October 2017

ER -

Guo J, Xiao B, Cai P , Qian W , Zhou A. Efficient multi-version storage engine for main memory data store. In Jia W, Klimenko A, Dzerzhinskiy F, Klimenko SV, Chen L, Li Q, Gao Y, Bouguettaya A, Zhang X, editors, Web Information Systems Engineering – WISE 2017 - 18th International Conference, Proceedings. Springer Verlag. 2017. p. 205-220. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-68786-5_17

Efficient multi-version storage engine for main memory data store

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this