Corrected trees for reliable group communication

Martin Küttler; Maksym Planeta; Jan Bierbaum; Carsten Weinhold; Hermann Härtig; Amnon Barak; Torsten Hoefler

doi:10.1145/3293883.3295721

Corrected trees for reliable group communication

Martin Küttler
, Maksym Planeta
, Jan Bierbaum
, Carsten Weinhold
, Hermann Härtig
, Amnon Barak
, Torsten Hoefler

The Rachel and Selim Benin School of Engineering and Computer Science

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

2 Scopus citations

Abstract

Driven by ever increasing performance demands of compute-intensive applications, supercomputing systems comprise more and more nodes. This growth is a significant burden for fast group communication primitives and also makes those systems more susceptible to failures of individual nodes. In this paper we present a two-phase fault-tolerant scheme for group communication. Using broadcast as an example, we provide a full-spectrum discussion of our approach - from a formal analysis to LogP-based simulations to a message-passing-based implementation running on a large cluster. Ultimately, we are able to reduce the complex problem of reliable and fault-tolerant collective group communication to a graph theoretical renumbering problem. Both, simulations and measurements, show our solution to achieve a latency reduction of 50% with up to six times fewer messages sent in comparison to existing schemes.

Original language	English
Title of host publication	PPoPP 2019 - Proceedings of the 24th Principles and Practice of Parallel Programming
Publisher	Association for Computing Machinery
Pages	287-299
Number of pages	13
ISBN (Electronic)	9781450362252
DOIs	https://doi.org/10.1145/3293883.3295721
State	Published - 16 Feb 2019
Event	24th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2019 - Washington, United States Duration: 16 Feb 2019 → 20 Feb 2019

Publication series

Name	Proceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP
ISSN (Print)	1542-0205

Conference

Conference	24th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2019
Country/Territory	United States
City	Washington
Period	16/02/19 → 20/02/19

Bibliographical note

Publisher Copyright:
© 2019 Copyright held by the owner/author(s).

Keywords

Gossip
HPC
LogP model
Low-latency broadcast
MPI

Access to Document

10.1145/3293883.3295721

Cite this

Küttler, M., Planeta, M., Bierbaum, J., Weinhold, C., Härtig, H., Barak, A., & Hoefler, T. (2019). Corrected trees for reliable group communication. In PPoPP 2019 - Proceedings of the 24th Principles and Practice of Parallel Programming (pp. 287-299). (Proceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP). Association for Computing Machinery. https://doi.org/10.1145/3293883.3295721

@inproceedings{dcc98853b09642a09619305b314cead4,

title = "Corrected trees for reliable group communication",

abstract = "Driven by ever increasing performance demands of compute-intensive applications, supercomputing systems comprise more and more nodes. This growth is a significant burden for fast group communication primitives and also makes those systems more susceptible to failures of individual nodes. In this paper we present a two-phase fault-tolerant scheme for group communication. Using broadcast as an example, we provide a full-spectrum discussion of our approach - from a formal analysis to LogP-based simulations to a message-passing-based implementation running on a large cluster. Ultimately, we are able to reduce the complex problem of reliable and fault-tolerant collective group communication to a graph theoretical renumbering problem. Both, simulations and measurements, show our solution to achieve a latency reduction of 50\% with up to six times fewer messages sent in comparison to existing schemes.",

keywords = "Gossip, HPC, LogP model, Low-latency broadcast, MPI",

author = "Martin K{\"u}ttler and Maksym Planeta and Jan Bierbaum and Carsten Weinhold and Hermann H{\"a}rtig and Amnon Barak and Torsten Hoefler",

note = "Publisher Copyright: {\textcopyright} 2019 Copyright held by the owner/author(s).; 24th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2019 ; Conference date: 16-02-2019 Through 20-02-2019",

year = "2019",

month = feb,

day = "16",

doi = "10.1145/3293883.3295721",

language = "אנגלית",

series = "Proceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP",

publisher = "Association for Computing Machinery",

pages = "287--299",

booktitle = "PPoPP 2019 - Proceedings of the 24th Principles and Practice of Parallel Programming",

}

Küttler, M, Planeta, M, Bierbaum, J, Weinhold, C, Härtig, H, Barak, A & Hoefler, T 2019, Corrected trees for reliable group communication. in PPoPP 2019 - Proceedings of the 24th Principles and Practice of Parallel Programming. Proceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP, Association for Computing Machinery, pp. 287-299, 24th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2019, Washington, United States, 16/02/19. https://doi.org/10.1145/3293883.3295721

Corrected trees for reliable group communication. / Küttler, Martin; Planeta, Maksym; Bierbaum, Jan et al.
PPoPP 2019 - Proceedings of the 24th Principles and Practice of Parallel Programming. Association for Computing Machinery, 2019. p. 287-299 (Proceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP).

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

TY - GEN

T1 - Corrected trees for reliable group communication

AU - Küttler, Martin

AU - Planeta, Maksym

AU - Bierbaum, Jan

AU - Weinhold, Carsten

AU - Härtig, Hermann

AU - Barak, Amnon

AU - Hoefler, Torsten

PY - 2019/2/16

Y1 - 2019/2/16

N2 - Driven by ever increasing performance demands of compute-intensive applications, supercomputing systems comprise more and more nodes. This growth is a significant burden for fast group communication primitives and also makes those systems more susceptible to failures of individual nodes. In this paper we present a two-phase fault-tolerant scheme for group communication. Using broadcast as an example, we provide a full-spectrum discussion of our approach - from a formal analysis to LogP-based simulations to a message-passing-based implementation running on a large cluster. Ultimately, we are able to reduce the complex problem of reliable and fault-tolerant collective group communication to a graph theoretical renumbering problem. Both, simulations and measurements, show our solution to achieve a latency reduction of 50% with up to six times fewer messages sent in comparison to existing schemes.

AB - Driven by ever increasing performance demands of compute-intensive applications, supercomputing systems comprise more and more nodes. This growth is a significant burden for fast group communication primitives and also makes those systems more susceptible to failures of individual nodes. In this paper we present a two-phase fault-tolerant scheme for group communication. Using broadcast as an example, we provide a full-spectrum discussion of our approach - from a formal analysis to LogP-based simulations to a message-passing-based implementation running on a large cluster. Ultimately, we are able to reduce the complex problem of reliable and fault-tolerant collective group communication to a graph theoretical renumbering problem. Both, simulations and measurements, show our solution to achieve a latency reduction of 50% with up to six times fewer messages sent in comparison to existing schemes.

KW - Gossip

KW - HPC

KW - LogP model

KW - Low-latency broadcast

KW - MPI

UR - https://www.scopus.com/pages/publications/85064232103

U2 - 10.1145/3293883.3295721

DO - 10.1145/3293883.3295721

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:85064232103

T3 - Proceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP

SP - 287

EP - 299

BT - PPoPP 2019 - Proceedings of the 24th Principles and Practice of Parallel Programming

PB - Association for Computing Machinery

T2 - 24th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2019

Y2 - 16 February 2019 through 20 February 2019

ER -

Küttler M, Planeta M, Bierbaum J, Weinhold C, Härtig H, Barak A et al. Corrected trees for reliable group communication. In PPoPP 2019 - Proceedings of the 24th Principles and Practice of Parallel Programming. Association for Computing Machinery. 2019. p. 287-299. (Proceedings of the ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPOPP). doi: 10.1145/3293883.3295721

Corrected trees for reliable group communication

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this