Cited By
View all- Abate AMarin A(2023)Introduction to the Special Issue on QEST 2021ACM Transactions on Modeling and Computer Simulation10.1145/363170733:4(1-2)Online publication date: 30-Nov-2023
Performance Analysis of Work Stealing Strategies in Large-Scale Multithreaded Computing
Article No.: 15, Pages 1 - 23
Abstract
Distributed systems use randomized work stealing to improve performance and resource utilization. In most prior analytical studies of randomized work stealing, jobs are considered to be sequential and are executed as a whole on a single server. In this article, we consider a homogeneous system of servers where parent jobs spawn child jobs that can feasibly be executed in parallel. When an idle server probes a busy server in an attempt to steal work, it may either steal a parent job or multiple child jobs.
To approximate the performance of this system, we introduce a Quasi-Birth-Death Markov chain and express the performance measures of interest via its unique steady state. We perform simulation experiments that suggest that the approximation error tends to zero as the number of servers in the system becomes large. To further support this observation, we introduce a mean field model and show that its unique fixed point corresponds to the steady state of the Quasi-Birth-Death Markov chain. Using numerical experiments, we compare the performance of various simple stealing strategies as well as optimized strategies.
References
[1]
Dario A. Bini, Beatrice Meini, Sergio Steffé, and Benny Van Houdt. 2006. Structured Markov chains solver: Software tools. In Proceeding of the 2006 Workshop on Tools for Solving Structured Markov Chains. 1–14.
[2]
M. Bladt and B. F. Nielsen. 2017. Matrix-Exponential Distributions in Applied Probability. Vol. 81. Springer.
[3]
Robert D. Blumofe, Christopher F. Joerg, Bradley C. Kuszmaul, Charles E. Leiserson, Keith H. Randall, and Yuli Zhou. 1996. Cilk: An efficient multithreaded runtime system. Journal of Parallel and Distributed Computing 37, 1 (1996), 55–69.
[4]
Robert D. Blumofe and Charles E. Leiserson. 1999. Scheduling multithreaded computations by work stealing. Journal of the ACM 46, 5 (1999), 720–748.
[5]
Maury Bramson, Yi Lu, and Balaji Prabhakar. 2010. Randomized load balancing with general service time distributions. In Proceedings of the ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems (SIGMETRICS’10). 275–286. DOI:
[6]
Derek L. Eager, Edward D. Lazowska, and John Zahorjan. 1986. A comparison of receiver-initiated and sender-initiated adaptive load sharing. Performance Evaluation 6, 1 (1986), 53–68.
[7]
Matteo Frigo, Charles E. Leiserson, and Keith H. Randall. 1998. The implementation of the Cilk-5 multithreaded language. In Proceedings of the ACM SIGPLAN 1998 Conference on Program Language Design and Implementation (PLDI’98). 212–223.
[8]
Nicolas Gast. 2017. Expected values estimated via mean-field approximation are 1/N-accurate. Proceedings of the ACM on Measurement and Analysis of Computing Systems 1, 1 (2017), Article 17, 26 pages.
[9]
Nicolas Gast and Bruno Gaujal. 2010. A mean field model of work stealing in large-scale systems. ACM SIGMETRICS Performance Evaluation Review 38, 1 (2010), 13–24.
[10]
Thierry Gautier, Xavier Besseron, and Laurent Pigeon. 2007. KAAPI: A thread scheduling runtime system for data flow computations on cluster of multi-processors. In Proceedings of the 2007 International Workshop on Parallel Symbolic Computation. 15–23.
[11]
Grzegorz Kielanski and Benny Van Houdt. 2021. Performance analysis of work stealing strategies in large scale multi-threaded computing. In Quantitative Evaluation of Systems. Springer International Publishing, Cham, Switzerland, 329–348.
[12]
J. Kriege and P. Buchholz. 2014. PH and MAP Fitting with Aggregated Traffic Traces. Springer International Publishing, Cham, Switzerland, 1–15. DOI:
[13]
Guy Latouche and V. Ramaswami. 1999. Introduction to Matrix Analytic Methods in Stochastic Modeling. Vol. 5. SIAM.
[14]
Doug Lea. 2000. A Java fork/join framework. In Proceedings of the ACM 2000 Conference on Java Grande (JAVA’00). ACM, New York, NY, 36–43.
[15]
Daan Leijen, Wolfram Schulte, and Sebastian Burckhardt. 2009. The design of a task parallel library. In Proceedings of the 24th ACM SIGPLAN Conference on Object Oriented Programming Systems Languages and Applications (OOPSLA’09). ACM, New York, NY, 227–242.
[16]
Wouter Minnebo, Tim Hellemans, and Benny Van Houdt. 2017. On a class of push and pull strategies with single migrations and limited probe rate. Performance Evaluation 113 (2017), 42–67.
[17]
Wouter Minnebo and Benny Van Houdt. 2014. A fair comparison of pull and push strategies in large distributed networks. IEEE/ACM Transactions on Networking 22, 3 (2014), 996–1006.
[18]
Ravi Mirchandaney, Don Towsley, and John A. Stankovic. 1990. Adaptive load sharing in heterogeneous distributed systems. Journal of Parallel and Distributed Computing 9, 4 (1990), 331–346.
[19]
Marcel F. Neuts. 1981. Matrix-Geometric Solutions in Stochastic Models: An Algorithmic Approach. John Hopkins University Press, Baltimore, MD.
[20]
A. Panchenko and A. Thümmler. 2007. Efficient phase-type fitting with aggregated traffic traces. Performance Evaluation 64, 7-8 (Aug.2007), 629–645. DOI:
[21]
Arch Robison, Michael Voss, and Alexey Kukanov. 2008. Optimization via reflection on work stealing in TBB. In Proceedings of the 2008 IEEE International Symposium on Parallel and Distributed Processing. IEEE, Los Alamitos, CA, 1–8.
[22]
Nikki Sonenberg, Grzegorz Kielanski, and Benny Van Houdt. 2021. Performance analysis of work stealing in large-scale multithreaded computing. ACM Transactions on Modeling and Performance Evaluation of Computing Systems 6, 2 (Sept. 2021), Article 6, 28 pages. DOI:
[23]
Ignace Van Spilbeeck and Benny Van Houdt. 2015. Performance of rate-based pull and push strategies in heterogeneous networks. Performance Evaluation 91 (2015), 2–15.
[24]
Mark S. Squillante and Randolph D. Nelson. 1991. Analysis of task migration in shared-memory multiprocessor scheduling. ACM SIGMETRICS Performance Evaluation Review 19, 1 (1991), 143–155.
[25]
Benny Van Houdt. 2019. Randomized work stealing versus sharing in large-scale systems with non-exponential job sizes. IEEE/ACM Transactions on Networking 27, 5 (2019), 2137–2149.
[26]
Niklaus Wirth. 1996. Tasks versus threads: An alternative multiprocessing paradigm. Softw. Concepts Tools 17 (1996), 6–12.
Index Terms
- Performance Analysis of Work Stealing Strategies in Large-Scale Multithreaded Computing
Recommendations
Performance Analysis of Work Stealing Strategies in Large Scale Multi-threaded Computing
Quantitative Evaluation of SystemsAbstractDistributed systems use randomized work stealing to improve performance and resource utilization. In most prior analytical studies of randomized work stealing, jobs are considered to be sequential and are executed as a whole on a single server. In ...
Performance Analysis of Work Stealing in Large-scale Multithreaded Computing
Randomized work stealing is used in distributed systems to increase performance and improve resource utilization. In this article, we consider randomized work stealing in a large system of homogeneous processors where parent jobs spawn child jobs that can ...
Adaptive work stealing with parallelism feedback
PPoPP '07: Proceedings of the 12th ACM SIGPLAN symposium on Principles and practice of parallel programmingWe present an adaptive work-stealing thread scheduler, A-Steal, for fork-join multithreaded jobs, like those written using the Cilk multithreaded language or the Hood work-stealing library. The A-Steal algorithm is appropriate for large parallel servers ...
Comments
Information & Contributors
Information
Published In
October 2023
175 pages
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 26 October 2023
Online AM: 16 February 2023
Accepted: 07 February 2023
Revised: 17 August 2022
Received: 19 January 2022
Published in TOMACS Volume 33, Issue 4
Check for updates
Author Tags
Qualifiers
- Research-article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- View Citations1Total Citations
- 150Total Downloads
- Downloads (Last 12 months)99
- Downloads (Last 6 weeks)13
Reflects downloads up to 10 Oct 2024
Other Metrics
Citations
Cited By
View all- Abate AMarin A(2023)Introduction to the Special Issue on QEST 2021ACM Transactions on Modeling and Computer Simulation10.1145/363170733:4(1-2)Online publication date: 30-Nov-2023
View Options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in