SwiSh: Distributed Shared State Abstractions for Programmable Switches

Lior Zeno; Dan R.K. Ports; Jacob Nelson; Daehyeok Kim; Shir Landau Feibish; Idit Keidar; Arik Rinberg; Alon Rashelbach; Igor De-Paula; Mark Silberstein

SwiSh: Distributed Shared State Abstractions for Programmable Switches

Lior Zeno
, Dan R.K. Ports
, Jacob Nelson
, Daehyeok Kim
, Shir Landau Feibish
, Idit Keidar
, Arik Rinberg
, Alon Rashelbach
, Igor De-Paula
, Mark Silberstein

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

Abstract

We design and evaluate SwiSh, a distributed shared state management layer for data-plane P4 programs. SwiSh enables running scalable stateful distributed network functions on programmable switches entirely in the data-plane. We explore several schemes to build a shared variable abstraction, which differ in consistency, performance, and in-switch implementation complexity. We introduce the novel Strong Delayed-Writes (SDW) protocol which offers consistent snapshots of shared data-plane objects with semantics known as r-relaxed strong linearizability, enabling implementation of distributed concurrent sketches with precise error bounds. We implement strong, eventual, and SDW consistency protocols in Tofino switches, and compare their performance in microbenchmarks and three realistic network functions, NAT, DDoS detector, and rate limiter. Our results show that the distributed state management in the data plane is practical, and outperforms centralized solutions by up to four orders of magnitude in update throughput and replication latency.

Original language	English
Title of host publication	Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022
Publisher	USENIX Association
Pages	171-191
Number of pages	21
ISBN (Electronic)	9781939133274
State	Published - 2022
Externally published	Yes
Event	19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022 - Renton, United States Duration: 4 Apr 2022 → 6 Apr 2022

Publication series

Name	Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022

Conference

Conference	19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022
Country/Territory	United States
City	Renton
Period	4/04/22 → 6/04/22

Bibliographical note

Publisher Copyright:
© 2022 by The USENIX Association. All Rights Reserved.

ASJC Scopus subject areas

Computer Networks and Communications
Control and Systems Engineering

Cite this

Zeno, L., Ports, D. R. K., Nelson, J., Kim, D., Feibish, S. L., Keidar, I., Rinberg, A., Rashelbach, A., De-Paula, I., & Silberstein, M. (2022). SwiSh: Distributed Shared State Abstractions for Programmable Switches. In Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022 (pp. 171-191). (Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022). USENIX Association.

@inproceedings{a61f0fe18b1844aea0739cc5a932af68,

title = "SwiSh: Distributed Shared State Abstractions for Programmable Switches",

abstract = "We design and evaluate SwiSh, a distributed shared state management layer for data-plane P4 programs. SwiSh enables running scalable stateful distributed network functions on programmable switches entirely in the data-plane. We explore several schemes to build a shared variable abstraction, which differ in consistency, performance, and in-switch implementation complexity. We introduce the novel Strong Delayed-Writes (SDW) protocol which offers consistent snapshots of shared data-plane objects with semantics known as r-relaxed strong linearizability, enabling implementation of distributed concurrent sketches with precise error bounds. We implement strong, eventual, and SDW consistency protocols in Tofino switches, and compare their performance in microbenchmarks and three realistic network functions, NAT, DDoS detector, and rate limiter. Our results show that the distributed state management in the data plane is practical, and outperforms centralized solutions by up to four orders of magnitude in update throughput and replication latency.",

author = "Lior Zeno and Ports, \{Dan R.K.\} and Jacob Nelson and Daehyeok Kim and Feibish, \{Shir Landau\} and Idit Keidar and Arik Rinberg and Alon Rashelbach and Igor De-Paula and Mark Silberstein",

note = "Publisher Copyright: {\textcopyright} 2022 by The USENIX Association. All Rights Reserved.; 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022 ; Conference date: 04-04-2022 Through 06-04-2022",

year = "2022",

language = "English",

series = "Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022",

publisher = "USENIX Association",

pages = "171--191",

booktitle = "Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022",

}

Zeno, L, Ports, DRK, Nelson, J, Kim, D, Feibish, SL, Keidar, I, Rinberg, A, Rashelbach, A, De-Paula, I & Silberstein, M 2022, SwiSh: Distributed Shared State Abstractions for Programmable Switches. in Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022. Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022, USENIX Association, pp. 171-191, 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022, Renton, United States, 4/04/22.

SwiSh: Distributed Shared State Abstractions for Programmable Switches. / Zeno, Lior; Ports, Dan R.K.; Nelson, Jacob et al.
Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022. USENIX Association, 2022. p. 171-191 (Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022).

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

TY - GEN

T1 - SwiSh

T2 - 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022

AU - Zeno, Lior

AU - Ports, Dan R.K.

AU - Nelson, Jacob

AU - Kim, Daehyeok

AU - Feibish, Shir Landau

AU - Keidar, Idit

AU - Rinberg, Arik

AU - Rashelbach, Alon

AU - De-Paula, Igor

AU - Silberstein, Mark

PY - 2022

Y1 - 2022

N2 - We design and evaluate SwiSh, a distributed shared state management layer for data-plane P4 programs. SwiSh enables running scalable stateful distributed network functions on programmable switches entirely in the data-plane. We explore several schemes to build a shared variable abstraction, which differ in consistency, performance, and in-switch implementation complexity. We introduce the novel Strong Delayed-Writes (SDW) protocol which offers consistent snapshots of shared data-plane objects with semantics known as r-relaxed strong linearizability, enabling implementation of distributed concurrent sketches with precise error bounds. We implement strong, eventual, and SDW consistency protocols in Tofino switches, and compare their performance in microbenchmarks and three realistic network functions, NAT, DDoS detector, and rate limiter. Our results show that the distributed state management in the data plane is practical, and outperforms centralized solutions by up to four orders of magnitude in update throughput and replication latency.

AB - We design and evaluate SwiSh, a distributed shared state management layer for data-plane P4 programs. SwiSh enables running scalable stateful distributed network functions on programmable switches entirely in the data-plane. We explore several schemes to build a shared variable abstraction, which differ in consistency, performance, and in-switch implementation complexity. We introduce the novel Strong Delayed-Writes (SDW) protocol which offers consistent snapshots of shared data-plane objects with semantics known as r-relaxed strong linearizability, enabling implementation of distributed concurrent sketches with precise error bounds. We implement strong, eventual, and SDW consistency protocols in Tofino switches, and compare their performance in microbenchmarks and three realistic network functions, NAT, DDoS detector, and rate limiter. Our results show that the distributed state management in the data plane is practical, and outperforms centralized solutions by up to four orders of magnitude in update throughput and replication latency.

UR - https://www.scopus.com/pages/publications/85140969409

M3 - Conference contribution

T3 - Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022

SP - 171

EP - 191

BT - Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022

PB - USENIX Association

Y2 - 4 April 2022 through 6 April 2022

ER -

Zeno L, Ports DRK, Nelson J, Kim D, Feibish SL, Keidar I et al. SwiSh: Distributed Shared State Abstractions for Programmable Switches. In Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022. USENIX Association. 2022. p. 171-191. (Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation, NSDI 2022).

SwiSh: Distributed Shared State Abstractions for Programmable Switches

Abstract

Publication series

Conference

Bibliographical note

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this