skip to main content
10.1145/3519270.3538443acmconferencesArticle/Chapter ViewAbstractPublication PagespodcConference Proceedingsconference-collections
research-article
Public Access

From Switch Scheduling to Datacenter Scheduling: Matching-Coordinated Greed is Good

Published: 21 July 2022 Publication History

Abstract

Packet scheduling over a switch (interconnect) fabric is a wellstudied problem in distributed computing, with known near-optimal distributed bipartite matching based protocols.
We initiate a theoretical study of distributed flow scheduling in datacenter networks. Building upon the observation that modern datacenter networks use Clos-like topologies similar to switch fabrics, we introduce a new k-sparse flow-matching (k=k-SFM) problem, a variant of the classical matching problem that captures the unique constraints imposed by flow scheduling in datacenter networks. In the k=k-SFM problem, we are given a weighted graph and an integer k. The goal is to assign a fractional flow value to each edge under the following three constraints: (1) for each edge, the assigned flow value is no greater than its input weight; (2) for each vertex, the sum of flow values assigned to edges incident to the vertex is at most the capacity of the vertex; and (3) for each vertex, at most k incident edges are assigned a non-zero flow value. The goal is to compute a feasible solution with the largest total fractional weight.

Supplementary Material

MP4 File (S6-T5.mp4)
presentation paper

References

[1]
M. Alizadeh, S. Yang, M. Sharif, S. Katti, N.McKeown, B. Prabhakar, and S. Shenker. pfabric: minimal near-optimal datacenter transport. In SIGCOMM, 2013.
[2]
T. E. Anderson, S. S. Owicki, J. B. Saxe, and C. P. Thacker. High speed switch scheduling for local area networks. ACM Transactions on Computer Systems, 1993.
[3]
B. Awerbuch, M. T. Hajiaghayi, R. D. Kleinberg, and T. Leighton. Online clientserver load balancing without global information. In SODA, 2005.
[4]
R. Bar-Yehuda, K. Censor-Hillel, M. Ghaffari, and G. Schwartzman. Distributed approximation of maximum independent set and maximum matching. In PODC, 2017.
[5]
S. J. Belongie, J. Malik, and J. Puzicha. Shape matching and object recognition using shape contexts. IEEE Trans. Pattern Anal. Mach. Intell., 24(4):509--522, 2002.
[6]
T. Benson, A. Akella, and D. A. Maltz. Network traffic characteristics of data centers in the wild. In IMC, 2010.
[7]
D. Bertsimas, V. F. Farias, and N. Trichakis. Fairness, efficiency, and flexibility in organ allocation for kidney transplantation. Oper. Res., 61(1):73--87, 2013.
[8]
G. E. Blelloch, J. T. Fineman, and J. Shun. Greedy sequential maximal independent set and matching are parallel on average. In SPAA, 2012.
[9]
Q. Cai, M. T. Arashloo, and R. Agarwal. dcPIM: Near-optimal proactive datacenter transport. In SIGCOMM, 2022.
[10]
Q. Cai, S. Chaudhary, M. Vuppalapati, J. Hwang, and R. Agarwal. Understanding host network stack overheads. In SIGCOMM, 2021.
[11]
L. Charlin and R. Zemel. The toronto paper matching system: an automated paper-reviewer assignment system. 2013.
[12]
S. Chattopadhyay, L. Higham, and K. Seyffarth. Dynamic and self-stabilizing distributed matching. In PODC, 2002.
[13]
T. Cheung. Graph traversal techniques and the maximum flow problem in distributed computation. IEEE Transactions on Software Engineering, 1983.
[14]
J. P. Dickerson and T. Sandholm. Futurematch: Combining human value judgments and machine learning to match in dynamic environments. In AAAI, 2015.
[15]
J. Drummond, A. Perrault, and F. Bacchus. SAT is an effective and complete method for solving stable matching problems with couples. In IJCAI, 2015.
[16]
R. Duan and S. Pettie. Linear-time approximation for maximum weight matching. Journal of the ACM, 2014.
[17]
J. Edmonds. Paths, trees, and flowers. Canadian Journal of mathematics, 1965.
[18]
J. R. Edmonds and E. L. Johnson. Matching: A well-solved class of integer linear programs. In Combinatorial Optimization - Eureka, You Shrink!, Papers Dedicated to Jack Edmonds, 5th International Workshop, volume 2570 of Lecture Notes in Computer Science, pages 27--30. Springer, 2001.
[19]
G. Even, M. Medina, and D. Ron. Distributed maximum matching in bounded degree graphs. In ICDCN, 2015.
[20]
N. Farrington, E. Rubow, and A. Vahdat. Data center switch architecture in the age of merchant silicon. In HOTI, 2009.
[21]
M. Fischer. Improved deterministic distributed matching via rounding. Distributed Computing, 2020.
[22]
P. X. Gao, A. Narayan, G. Kumar, R. Agarwal, S. Ratnasamy, and S. Shenker. phost: distributed near-optimal datacenter transport over commodity network fabric. In CoNext, 2015.
[23]
M. Ghaffari, T. Gouleakis, C. Konrad, S. Mitrovic, and R. Rubinfeld. Improved massively parallel computation algorithms for mis, matching, and vertex cover. In PODC, 2018.
[24]
M. Ghaffari, A. Karrenbauer, F. Kuhn, C. Lenzen, and B. Patt-Shamir. Near-optimal distributed maximum flow: Extended abstract. In PODC, 2015.
[25]
S. Hu, W. Bai, G. Zeng, Z. Wang, B. Qiao, K. Chen, K. Tan, and Y. Wang. Aeolus: A building block for proactive transport in datacenters. In H. Schulzrinne and V. Misra, editors, SIGCOMM, 2020.
[26]
A. Israeli and A. Itai. A fast and simple randomized parallel algorithm for maximal matching. Information Processing Letters, 1986.
[27]
A. M. Khan, A. Pothen, M. M. A. Patwary, M. Halappanavar, N. R. Satish, N. Sundaram, and P. Dubey. Designing scalable b-matching algorithms on distributed memory multiprocessors by approximation. In SC, 2016.
[28]
C. Koufogiannakis and N. E. Young. Distributed fractional packing and maximum weighted b-matching via tail-recursive duality. In DISC, 2009.
[29]
E. Krissinel and K. Henrick. Secondary-structure matching (ssm), a new tool for fast protein structure alignment in three dimensions. Acta Crystallographica Section D: Biological Crystallography, 2004.
[30]
F. Kuhn, T. Moscibroda, and R. Wattenhofer. Local computation: Lower and upper bounds. Journal of the ACM, 2016.
[31]
R. Kurata, N. Hamada, A. Iwasaki, and M. Yokoo. Controlled school choice with soft bounds and overlapping types. J. Artif. Intell. Res., 58:153--184, 2017.
[32]
S. Lattanzi, B. Moseley, S. Suri, and S. Vassilvitskii. Filtering: a method for solving graph problems in mapreduce. In SPAA, 2011.
[33]
N. Linial. Distributive graph algorithms-global solutions from local data. In FOCS, 1987.
[34]
X. Liu, T. Suel, and N. D. Memon. A robust model for paper reviewer assignment. In RecSys, 2014.
[35]
Z. Lotker, B. Patt-Shamir, and S. Pettie. Improved distributed approximate matching. In SPAA, 2008.
[36]
Z. Lotker, B. Patt-Shamir, and S. Pettie. Improved distributed approximate matching. Journal of the ACM, 2015.
[37]
M. A. Marsan, A. Bianco, P. Giaccone, E. Leonardi, and F. Neri. Multicast traffic in input-queued switches: optimal scheduling and maximum throughput. IEEE/ACM Transactions on Networking, 2003.
[38]
N. McKeown. The iSLIP scheduling algorithm for input-queued switches. IEEE/ACM Transactions on Networking, 1999.
[39]
N. McKeown, A. Mekkittikul, V. Anantharam, and J. C. Walrand. Achieving 100% throughput in an input-queued switch. IEEE Transactions on Communications, 1999.
[40]
N. McKeown, P. Varaiya, and J. Walrand. Scheduling cells in an input-queued switch. Electronics Letters, 1993.
[41]
B. Montazeri, Y. Li, M. Alizadeh, and J. K. Ousterhout. Homa: a receiver-driven low-latency transport protocol using network priorities. In SIGCOMM, 2018.
[42]
A. Panconesi and M. Sozio. Fast primal-dual distributed algorithms for scheduling and matching problems. Distributed Computing, 2010.
[43]
J. Perry, A. Ousterhout, H. Balakrishnan, D. Shah, and H. Fugal. Fastpass: a centralized "zero-queue" datacenter network. In SIGCOMM, 2014.
[44]
T. L. Pham, I. Lavallée, M. Bui, and S. H. Do. A distributed algorithm for the maximum flow problem. In (ISPDC), 2005.
[45]
M. D. Plummer and L. Lovász. Matching theory. Elsevier, 1986.
[46]
W. R. Pulleyblank. Matchings and extensions. Handbook of combinatorics, 1:179-- 232, 1995.
[47]
A. Roy, H. Zeng, J. Bagga, G. Porter, and A. C. Snoeren. Inside the social network's (datacenter) network. In SIGCOMM, 2015.
[48]
D. Shah, P. Giaccone, and B. Prabhakar. Efficient randomized algorithms for input-queued switch scheduling. IEEE Micro, 2002.
[49]
D. Shah and D. Wischik. Optimal scheduling algorithms for input-queued switches. In INFOCOM, 2006.
[50]
R. Shashidhara, T. Stamler, A. Kaufmann, and S. Peter. FlexTOE: Flexible TCP offload with fine-grained parallelism. In NSDI, 2022.
[51]
A. Singh, J. Ong, A. Agarwal, G. Anderson, A. Armistead, R. Bannon, S. Boving, G. Desai, B. Felderman, P. Germano, A. Kanagala, J. Provost, J. Simmons, E. Tanda, J. Wanderer, U. Hölzle, S. Stuart, and A. Vahdat. Jupiter rising: A decade of clos topologies and centralized control in google's datacenter network. In SIGCOMM, 2015.
[52]
R. E. Tarjan. Data structures and network algorithms, volume 44 of CBMS-NSF regional conference series in applied mathematics. SIAM, 1983.
[53]
M. Wattenhofer and R. Wattenhofer. Distributed weighted matching. In DISC, 2004.

Cited By

View all
  • (2023)Opportunistic CPU Sharing in Mobile Edge Computing Deploying the Cloud-RANIEEE Transactions on Network and Service Management10.1109/TNSM.2023.330406720:3(2201-2217)Online publication date: 10-Aug-2023

Recommendations

Comments

Information & Contributors

Information

Published In

cover image ACM Conferences
PODC'22: Proceedings of the 2022 ACM Symposium on Principles of Distributed Computing
July 2022
509 pages
ISBN:9781450392624
DOI:10.1145/3519270
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].

Sponsors

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 21 July 2022

Permissions

Request permissions for this article.

Check for updates

Author Tags

  1. datacenter networks
  2. distributed matching algorithms
  3. flow scheduling protocols

Qualifiers

  • Research-article

Funding Sources

Conference

PODC '22
Sponsor:

Acceptance Rates

Overall Acceptance Rate 740 of 2,477 submissions, 30%

Contributors

Other Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

  • Downloads (Last 12 months)80
  • Downloads (Last 6 weeks)21
Reflects downloads up to 17 Feb 2025

Other Metrics

Citations

Cited By

View all
  • (2023)Opportunistic CPU Sharing in Mobile Edge Computing Deploying the Cloud-RANIEEE Transactions on Network and Service Management10.1109/TNSM.2023.330406720:3(2201-2217)Online publication date: 10-Aug-2023

View Options

View options

PDF

View or Download as a PDF file.

PDF

eReader

View online with eReader.

eReader

Login options

Figures

Tables

Media

Share

Share

Share this Publication link

Share on social media