A Highly Scalable Index Structure for Multicore In-Memory Database Systems

Hitoshi Mitake; Hiroshi Yamada; Tatsuo Nakajima

doi:10.1007/978-3-030-32258-8_25

A Highly Scalable Index Structure for Multicore In-Memory Database Systems

Hitoshi Mitake, Hiroshi Yamada, Tatsuo Nakajima^*

^*Corresponding author for this work

School of Fundamental Science and Engineering

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

Abstract

In this paper, we present some insights from the analysis of the drawbacks of the advanced concurrency control techniques. Based on the analysis, we reveal that a commonly used technique for index structures, read-copy update (RCU), has the most significant impact on the throughput and latency of in-memory database systems. For overcoming the drawbacks, we developed Glasstree, a new index structure that produces a smaller load on the memory allocator by enhancing Masstree. Glasstree achieved high throughput and stable latency than Masstree under various workloads.

Original language	English
Title of host publication	Intelligent Distributed Computing XIII, IDC 2019
Editors	Igor Kotenko, Vasily Desnitsky, Costin Badica, Didier El Baz, Mirjana Ivanovic
Publisher	Springer
Pages	210-217
Number of pages	8
ISBN (Print)	9783030322571
DOIs	https://doi.org/10.1007/978-3-030-32258-8_25
Publication status	Published - 2020
Event	13th International Symposium on Intelligent Distributed Computing, IDC 2019 - St. Petersburg, Russian Federation Duration: 2019 Oct 7 → 2019 Oct 9

Publication series

Name	Studies in Computational Intelligence
Volume	868
ISSN (Print)	1860-949X
ISSN (Electronic)	1860-9503

Conference

Conference	13th International Symposium on Intelligent Distributed Computing, IDC 2019
Country/Territory	Russian Federation
City	St. Petersburg
Period	19/10/7 → 19/10/9

Keywords

Epoch-based reclamation
In-memory database
Index tree structure
Multicore scalability

ASJC Scopus subject areas

Artificial Intelligence

Access to Document

10.1007/978-3-030-32258-8_25

Cite this

A Highly Scalable Index Structure for Multicore In-Memory Database Systems. / Mitake, Hitoshi; Yamada, Hiroshi; Nakajima, Tatsuo.

Intelligent Distributed Computing XIII, IDC 2019. ed. / Igor Kotenko; Vasily Desnitsky; Costin Badica; Didier El Baz; Mirjana Ivanovic. Springer, 2020. p. 210-217 (Studies in Computational Intelligence; Vol. 868).

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

Mitake, H, Yamada, H & Nakajima, T 2020, A Highly Scalable Index Structure for Multicore In-Memory Database Systems. in I Kotenko, V Desnitsky, C Badica, D El Baz & M Ivanovic (eds), Intelligent Distributed Computing XIII, IDC 2019. Studies in Computational Intelligence, vol. 868, Springer, pp. 210-217, 13th International Symposium on Intelligent Distributed Computing, IDC 2019, St. Petersburg, Russian Federation, 19/10/7. https://doi.org/10.1007/978-3-030-32258-8_25

@inproceedings{561d4a92bc2c4b34a531a3938c0819fd,

title = "A Highly Scalable Index Structure for Multicore In-Memory Database Systems",

abstract = "In this paper, we present some insights from the analysis of the drawbacks of the advanced concurrency control techniques. Based on the analysis, we reveal that a commonly used technique for index structures, read-copy update (RCU), has the most significant impact on the throughput and latency of in-memory database systems. For overcoming the drawbacks, we developed Glasstree, a new index structure that produces a smaller load on the memory allocator by enhancing Masstree. Glasstree achieved high throughput and stable latency than Masstree under various workloads.",

keywords = "Epoch-based reclamation, In-memory database, Index tree structure, Multicore scalability",

author = "Hitoshi Mitake and Hiroshi Yamada and Tatsuo Nakajima",

year = "2020",

doi = "10.1007/978-3-030-32258-8_25",

language = "English",

isbn = "9783030322571",

series = "Studies in Computational Intelligence",

publisher = "Springer",

pages = "210--217",

editor = "Igor Kotenko and Vasily Desnitsky and Costin Badica and {El Baz}, Didier and Mirjana Ivanovic",

booktitle = "Intelligent Distributed Computing XIII, IDC 2019",

note = "13th International Symposium on Intelligent Distributed Computing, IDC 2019 ; Conference date: 07-10-2019 Through 09-10-2019",

}

TY - GEN

T1 - A Highly Scalable Index Structure for Multicore In-Memory Database Systems

AU - Mitake, Hitoshi

AU - Yamada, Hiroshi

AU - Nakajima, Tatsuo

PY - 2020

Y1 - 2020

N2 - In this paper, we present some insights from the analysis of the drawbacks of the advanced concurrency control techniques. Based on the analysis, we reveal that a commonly used technique for index structures, read-copy update (RCU), has the most significant impact on the throughput and latency of in-memory database systems. For overcoming the drawbacks, we developed Glasstree, a new index structure that produces a smaller load on the memory allocator by enhancing Masstree. Glasstree achieved high throughput and stable latency than Masstree under various workloads.

AB - In this paper, we present some insights from the analysis of the drawbacks of the advanced concurrency control techniques. Based on the analysis, we reveal that a commonly used technique for index structures, read-copy update (RCU), has the most significant impact on the throughput and latency of in-memory database systems. For overcoming the drawbacks, we developed Glasstree, a new index structure that produces a smaller load on the memory allocator by enhancing Masstree. Glasstree achieved high throughput and stable latency than Masstree under various workloads.

KW - Epoch-based reclamation

KW - In-memory database

KW - Index tree structure

KW - Multicore scalability

UR - http://www.scopus.com/inward/record.url?scp=85075582812&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85075582812&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-32258-8_25

DO - 10.1007/978-3-030-32258-8_25

M3 - Conference contribution

AN - SCOPUS:85075582812

SN - 9783030322571

T3 - Studies in Computational Intelligence

SP - 210

EP - 217

BT - Intelligent Distributed Computing XIII, IDC 2019

A2 - Kotenko, Igor

A2 - Desnitsky, Vasily

A2 - Badica, Costin

A2 - El Baz, Didier

A2 - Ivanovic, Mirjana

PB - Springer

T2 - 13th International Symposium on Intelligent Distributed Computing, IDC 2019

Y2 - 7 October 2019 through 9 October 2019

ER -

A Highly Scalable Index Structure for Multicore In-Memory Database Systems

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this