Strong and efficient consistency with consistency-aware durability

Aishwarya Ganesan; Ramnatthan Alagappan; Andrea C. Arpaci-Dusseau; Remzi H. Arpaci-Dusseau

Strong and efficient consistency with consistency-aware durability

Aishwarya Ganesan
, Ramnatthan Alagappan
, Andrea C. Arpaci-Dusseau
, Remzi H. Arpaci-Dusseau

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

Abstract

We introduce consistency-aware durability or CAD, a new approach to durability in distributed storage that enables strong consistency while delivering high performance. We demonstrate the efficacy of this approach by designing cross-client monotonic reads, a novel and strong consistency property that provides monotonic reads across failures and sessions in leader-based systems. We build ORCA, a modified version of ZooKeeper that implements CAD and cross-client monotonic reads. We experimentally show that ORCA provides strong consistency while closely matching the performance of weakly consistent ZooKeeper. Compared to strongly consistent ZooKeeper, ORCA provides significantly higher throughput (1.8 - 3.3×), and notably reduces latency, sometimes by an order of magnitude in geo-distributed settings.

Original language	English (US)
Title of host publication	Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020
Publisher	USENIX Association
Pages	323-337
Number of pages	15
ISBN (Electronic)	9781939133120
State	Published - 2020
Externally published	Yes
Event	18th USENIX Conference on File and Storage Technologies, FAST 2020 - Santa Clara, United States Duration: Feb 25 2020 → Feb 27 2020

Conference

Conference	18th USENIX Conference on File and Storage Technologies, FAST 2020
Country/Territory	United States
City	Santa Clara
Period	2/25/20 → 2/27/20

ASJC Scopus subject areas

Hardware and Architecture
Software
Computer Networks and Communications

Library availability

Discover UIUC Full Text

Cite this

@inproceedings{80d5afa5f479466b83dffa8910f1a359,

title = "Strong and efficient consistency with consistency-aware durability",

abstract = "We introduce consistency-aware durability or CAD, a new approach to durability in distributed storage that enables strong consistency while delivering high performance. We demonstrate the efficacy of this approach by designing cross-client monotonic reads, a novel and strong consistency property that provides monotonic reads across failures and sessions in leader-based systems. We build ORCA, a modified version of ZooKeeper that implements CAD and cross-client monotonic reads. We experimentally show that ORCA provides strong consistency while closely matching the performance of weakly consistent ZooKeeper. Compared to strongly consistent ZooKeeper, ORCA provides significantly higher throughput (1.8 - 3.3×), and notably reduces latency, sometimes by an order of magnitude in geo-distributed settings.",

author = "Aishwarya Ganesan and Ramnatthan Alagappan and Arpaci-Dusseau, \{Andrea C.\} and Arpaci-Dusseau, \{Remzi H.\}",

note = "Funding Information: We thank Yu Hua (our shepherd) and the anonymous reviewers of FAST {\textquoteright}20 for their insightful comments and suggestions. We thank the members of ADSL for their excellent feedback. We also thank CloudLab [47] for providing a great environment to run our experiments. This material was supported by funding from NSF grants CNS-1421033, CNS-1763810 and CNS-1838733, and DOE grant DE-SC0014935. Aishwarya Ganesan is supported by a Facebook fellowship. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and may not reflect the views of NSF, DOE, or any other institutions. Publisher Copyright: Copyright {\textcopyright} Proc. of the 18th USENIX Conference on File and Storage Tech., FAST 2020. All rights reserved.; 18th USENIX Conference on File and Storage Technologies, FAST 2020 ; Conference date: 25-02-2020 Through 27-02-2020",

year = "2020",

language = "English (US)",

pages = "323--337",

booktitle = "Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020",

publisher = "USENIX Association",

}

TY - GEN

T1 - Strong and efficient consistency with consistency-aware durability

AU - Ganesan, Aishwarya

AU - Alagappan, Ramnatthan

AU - Arpaci-Dusseau, Andrea C.

AU - Arpaci-Dusseau, Remzi H.

N1 - Funding Information: We thank Yu Hua (our shepherd) and the anonymous reviewers of FAST ’20 for their insightful comments and suggestions. We thank the members of ADSL for their excellent feedback. We also thank CloudLab [47] for providing a great environment to run our experiments. This material was supported by funding from NSF grants CNS-1421033, CNS-1763810 and CNS-1838733, and DOE grant DE-SC0014935. Aishwarya Ganesan is supported by a Facebook fellowship. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and may not reflect the views of NSF, DOE, or any other institutions. Publisher Copyright: Copyright © Proc. of the 18th USENIX Conference on File and Storage Tech., FAST 2020. All rights reserved.

PY - 2020

Y1 - 2020

N2 - We introduce consistency-aware durability or CAD, a new approach to durability in distributed storage that enables strong consistency while delivering high performance. We demonstrate the efficacy of this approach by designing cross-client monotonic reads, a novel and strong consistency property that provides monotonic reads across failures and sessions in leader-based systems. We build ORCA, a modified version of ZooKeeper that implements CAD and cross-client monotonic reads. We experimentally show that ORCA provides strong consistency while closely matching the performance of weakly consistent ZooKeeper. Compared to strongly consistent ZooKeeper, ORCA provides significantly higher throughput (1.8 - 3.3×), and notably reduces latency, sometimes by an order of magnitude in geo-distributed settings.

AB - We introduce consistency-aware durability or CAD, a new approach to durability in distributed storage that enables strong consistency while delivering high performance. We demonstrate the efficacy of this approach by designing cross-client monotonic reads, a novel and strong consistency property that provides monotonic reads across failures and sessions in leader-based systems. We build ORCA, a modified version of ZooKeeper that implements CAD and cross-client monotonic reads. We experimentally show that ORCA provides strong consistency while closely matching the performance of weakly consistent ZooKeeper. Compared to strongly consistent ZooKeeper, ORCA provides significantly higher throughput (1.8 - 3.3×), and notably reduces latency, sometimes by an order of magnitude in geo-distributed settings.

UR - https://www.scopus.com/pages/publications/85091819502

UR - https://www.scopus.com/pages/publications/85091819502#tab=citedBy

M3 - Conference contribution

AN - SCOPUS:85091819502

SP - 323

EP - 337

BT - Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020

PB - USENIX Association

T2 - 18th USENIX Conference on File and Storage Technologies, FAST 2020

Y2 - 25 February 2020 through 27 February 2020

ER -

Strong and efficient consistency with consistency-aware durability

Abstract

Conference

ASJC Scopus subject areas

Library availability

Related links

Fingerprint

Cite this