Charon: Load-Aware Load-Balancing in P4

Carmine Rizzi; Zhiyuan Yao; Yoann Desmouceaux; Mark Townsley; Thomas Clausen

doi:10.23919/CNSM52442.2021.9615535

Charon: Load-Aware Load-Balancing in P4

Carmine Rizzi
, Zhiyuan Yao
, Yoann Desmouceaux
, Mark Townsley
, Thomas Clausen

École Polytechnique

Cisco Systems

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

Abstract

Load-Balancers play an important role in data centers as they distribute network flows across application servers and guarantee per-connection consistency. It is hard however to make fair load balancing decisions so that all resources are efficiently occupied yet not overloaded. Tracking connection states allows to infer server load states and make informed decisions, but at the cost of additional memory space consumption. This makes it hard to implement on programmable hardware, which has constrained memory but offers line-rate performance. This paper presents Charon, a stateless load-aware load balancer that has line-rate performance implemented in P4-NetFPGA. Charon passively collects load states from application servers and employs the power-of-2-choices scheme to make data-driven load balancing decisions and improve resource utilization. Per-connection consistency is preserved statelessly by encoding server ID in a covert channel. The prototype design and implementation details are described in this paper. Simulation results show performance gains in terms of load distribution fairness, quality of service, throughput and processing latency.

Original language	English
Title of host publication	Proceedings of the 2021 17th International Conference on Network and Service Management
Subtitle of host publication	Smart Management for Future Networks and Services, CNSM 2021
Editors	Prosper Chemouil, Mehmet Ulema, Stuart Clayman, Muge Sayit, Cihat Cetinkaya, Stefano Secci
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	91-97
Number of pages	7
ISBN (Electronic)	9783903176362
DOIs	https://doi.org/10.23919/CNSM52442.2021.9615535
Publication status	Published - 1 Jan 2021
Event	17th International Conference on Network and Service Management, CNSM 2021 - Virtual, Online, Turkey Duration: 25 Oct 2021 → 29 Oct 2021

Publication series

Name	Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021

Conference

Conference	17th International Conference on Network and Service Management, CNSM 2021
Country/Territory	Turkey
City	Virtual, Online
Period	25/10/21 → 29/10/21

Keywords

cloud and distributed computing
load-balancing

Access to Document

10.23919/CNSM52442.2021.9615535

Cite this

Rizzi, C., Yao, Z., Desmouceaux, Y., Townsley, M., & Clausen, T. (2021). Charon: Load-Aware Load-Balancing in P4. In P. Chemouil, M. Ulema, S. Clayman, M. Sayit, C. Cetinkaya, & S. Secci (Eds.), Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021 (pp. 91-97). (Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.23919/CNSM52442.2021.9615535

Rizzi, Carmine ; Yao, Zhiyuan ; Desmouceaux, Yoann et al. / Charon : Load-Aware Load-Balancing in P4. Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021. editor / Prosper Chemouil ; Mehmet Ulema ; Stuart Clayman ; Muge Sayit ; Cihat Cetinkaya ; Stefano Secci. Institute of Electrical and Electronics Engineers Inc., 2021. pp. 91-97 (Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021).

@inproceedings{d97dae67c6964654a9af4dfde1ea3bf6,

title = "Charon: Load-Aware Load-Balancing in P4",

abstract = "Load-Balancers play an important role in data centers as they distribute network flows across application servers and guarantee per-connection consistency. It is hard however to make fair load balancing decisions so that all resources are efficiently occupied yet not overloaded. Tracking connection states allows to infer server load states and make informed decisions, but at the cost of additional memory space consumption. This makes it hard to implement on programmable hardware, which has constrained memory but offers line-rate performance. This paper presents Charon, a stateless load-aware load balancer that has line-rate performance implemented in P4-NetFPGA. Charon passively collects load states from application servers and employs the power-of-2-choices scheme to make data-driven load balancing decisions and improve resource utilization. Per-connection consistency is preserved statelessly by encoding server ID in a covert channel. The prototype design and implementation details are described in this paper. Simulation results show performance gains in terms of load distribution fairness, quality of service, throughput and processing latency.",

keywords = "cloud and distributed computing, load-balancing",

author = "Carmine Rizzi and Zhiyuan Yao and Yoann Desmouceaux and Mark Townsley and Thomas Clausen",

note = "Publisher Copyright: {\textcopyright} 2021 IFIP.; 17th International Conference on Network and Service Management, CNSM 2021 ; Conference date: 25-10-2021 Through 29-10-2021",

year = "2021",

month = jan,

day = "1",

doi = "10.23919/CNSM52442.2021.9615535",

language = "English",

series = "Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "91--97",

editor = "Prosper Chemouil and Mehmet Ulema and Stuart Clayman and Muge Sayit and Cihat Cetinkaya and Stefano Secci",

booktitle = "Proceedings of the 2021 17th International Conference on Network and Service Management",

}

Rizzi, C, Yao, Z, Desmouceaux, Y, Townsley, M & Clausen, T 2021, Charon: Load-Aware Load-Balancing in P4. in P Chemouil, M Ulema, S Clayman, M Sayit, C Cetinkaya & S Secci (eds), Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021. Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021, Institute of Electrical and Electronics Engineers Inc., pp. 91-97, 17th International Conference on Network and Service Management, CNSM 2021, Virtual, Online, Turkey, 25/10/21. https://doi.org/10.23919/CNSM52442.2021.9615535

Charon: Load-Aware Load-Balancing in P4. / Rizzi, Carmine; Yao, Zhiyuan; Desmouceaux, Yoann et al.
Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021. ed. / Prosper Chemouil; Mehmet Ulema; Stuart Clayman; Muge Sayit; Cihat Cetinkaya; Stefano Secci. Institute of Electrical and Electronics Engineers Inc., 2021. p. 91-97 (Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021).

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

TY - GEN

T1 - Charon

T2 - 17th International Conference on Network and Service Management, CNSM 2021

AU - Rizzi, Carmine

AU - Yao, Zhiyuan

AU - Desmouceaux, Yoann

AU - Townsley, Mark

AU - Clausen, Thomas

PY - 2021/1/1

Y1 - 2021/1/1

N2 - Load-Balancers play an important role in data centers as they distribute network flows across application servers and guarantee per-connection consistency. It is hard however to make fair load balancing decisions so that all resources are efficiently occupied yet not overloaded. Tracking connection states allows to infer server load states and make informed decisions, but at the cost of additional memory space consumption. This makes it hard to implement on programmable hardware, which has constrained memory but offers line-rate performance. This paper presents Charon, a stateless load-aware load balancer that has line-rate performance implemented in P4-NetFPGA. Charon passively collects load states from application servers and employs the power-of-2-choices scheme to make data-driven load balancing decisions and improve resource utilization. Per-connection consistency is preserved statelessly by encoding server ID in a covert channel. The prototype design and implementation details are described in this paper. Simulation results show performance gains in terms of load distribution fairness, quality of service, throughput and processing latency.

AB - Load-Balancers play an important role in data centers as they distribute network flows across application servers and guarantee per-connection consistency. It is hard however to make fair load balancing decisions so that all resources are efficiently occupied yet not overloaded. Tracking connection states allows to infer server load states and make informed decisions, but at the cost of additional memory space consumption. This makes it hard to implement on programmable hardware, which has constrained memory but offers line-rate performance. This paper presents Charon, a stateless load-aware load balancer that has line-rate performance implemented in P4-NetFPGA. Charon passively collects load states from application servers and employs the power-of-2-choices scheme to make data-driven load balancing decisions and improve resource utilization. Per-connection consistency is preserved statelessly by encoding server ID in a covert channel. The prototype design and implementation details are described in this paper. Simulation results show performance gains in terms of load distribution fairness, quality of service, throughput and processing latency.

KW - cloud and distributed computing

KW - load-balancing

UR - https://www.scopus.com/pages/publications/85123395337

U2 - 10.23919/CNSM52442.2021.9615535

DO - 10.23919/CNSM52442.2021.9615535

M3 - Conference contribution

AN - SCOPUS:85123395337

T3 - Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021

SP - 91

EP - 97

BT - Proceedings of the 2021 17th International Conference on Network and Service Management

A2 - Chemouil, Prosper

A2 - Ulema, Mehmet

A2 - Clayman, Stuart

A2 - Sayit, Muge

A2 - Cetinkaya, Cihat

A2 - Secci, Stefano

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 25 October 2021 through 29 October 2021

ER -

Rizzi C, Yao Z, Desmouceaux Y, Townsley M, Clausen T. Charon: Load-Aware Load-Balancing in P4. In Chemouil P, Ulema M, Clayman S, Sayit M, Cetinkaya C, Secci S, editors, Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021. Institute of Electrical and Electronics Engineers Inc. 2021. p. 91-97. (Proceedings of the 2021 17th International Conference on Network and Service Management: Smart Management for Future Networks and Services, CNSM 2021). doi: 10.23919/CNSM52442.2021.9615535

Charon: Load-Aware Load-Balancing in P4

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this