ZTE

ZTE Communications ›› 2014, Vol. 12 ›› Issue (2): 57-62.DOI: DOI:10.3969/j.issn.1673-5188.2014.02.009

• Special Topic • Previous Articles     Next Articles

Load Balancing Fat-Tree on Long-Lived Flows: Avoiding Congestion in a Data Center Network

Wen Gao, Xuyan Li, Boyang Zhou, and Chunming Wu   

  1. College of Computer Science, Zhejiang University, Hangzhou 310027, China
  • Received:2014-03-03 Online:2014-06-25 Published:2014-06-25
  • About author:Wen Gao (gavingao@zju.edu.cn) is a PhD candidate at the New Generation Network Technology Laboratory at Zhejiang University. His current research interests include network management, software-defined networks, and reconfigurable flexible networks.

    Xuyan Li (lixuyanzju@163.com) received her BS degree from Zhejiang University in 2014. Her research interests include software-defined networking and data center networking.

    Boyang Zhou (zby@zju.edu.cn ) is a PhD candidate at the New Generation Network Technology Laboratory at Zhejiang University. His current research interests include future Internet architecture, software-defined networks, and reconfigurable flexible networks.

    Chunming Wu (wuchunming@zju.edu.cn) is a professor in the College of Computer Science at Zhejiang University. He is the director of the New Generation Network Technology Laboratory at that university. His research interests include flexible reconfigurable networks, software-defined networks, network and service testbeds, and innovative security technology for active defense networks.
  • Supported by:
    This work is supported by the National Basic Research Program of China (973 Program) (2012CB315903), the Key Science and Technology Innovation Team Project of Zhejiang Province (2011R50010-05), the National Science and Technology Support Program (2014BAH24F01), 863 Program of China (2012AA01A507), and the National Natural Science Foundation of China (61379118 and 61103200). This work is sponsored by the Research Fund of ZTE Corporation.

Load Balancing Fat-Tree on Long-Lived Flows: Avoiding Congestion in a Data Center Network

Wen Gao, Xuyan Li, Boyang Zhou, and Chunming Wu   

  1. College of Computer Science, Zhejiang University, Hangzhou 310027, China
  • 作者简介:Wen Gao (gavingao@zju.edu.cn) is a PhD candidate at the New Generation Network Technology Laboratory at Zhejiang University. His current research interests include network management, software-defined networks, and reconfigurable flexible networks.

    Xuyan Li (lixuyanzju@163.com) received her BS degree from Zhejiang University in 2014. Her research interests include software-defined networking and data center networking.

    Boyang Zhou (zby@zju.edu.cn ) is a PhD candidate at the New Generation Network Technology Laboratory at Zhejiang University. His current research interests include future Internet architecture, software-defined networks, and reconfigurable flexible networks.

    Chunming Wu (wuchunming@zju.edu.cn) is a professor in the College of Computer Science at Zhejiang University. He is the director of the New Generation Network Technology Laboratory at that university. His research interests include flexible reconfigurable networks, software-defined networks, network and service testbeds, and innovative security technology for active defense networks.
  • 基金资助:
    This work is supported by the National Basic Research Program of China (973 Program) (2012CB315903), the Key Science and Technology Innovation Team Project of Zhejiang Province (2011R50010-05), the National Science and Technology Support Program (2014BAH24F01), 863 Program of China (2012AA01A507), and the National Natural Science Foundation of China (61379118 and 61103200). This work is sponsored by the Research Fund of ZTE Corporation.

PDF

Abstract: In a data center network (DCN), load balancing is required when servers transfer data on the same path. This is necessary to avoid congestion. Load balancing is challenged by the dynamic transferral of demands and complex routing control. Because of the distributed nature of a traditional network, previous research on load balancing has mostly focused on improving the performance of the local network; thus, the load has not been optimally balanced across the entire network. In this paper, we propose a novel dynamic load-balancing algorithm for fat-tree. This algorithm avoids congestions to the great possible extent by searching for non-conflicting paths in a centralized way. We implement the algorithm in the popular software-defined networking architecture and evaluate the algorithm’s performance on the Mininet platform. The results show that our algorithm has higher bisection bandwidth than the traditional equal-cost multi-path load-balancing algorithm and thus more effectively avoids congestion.

Key words: data center network, software-defined networking, load balancing, network management

摘要： In a data center network (DCN), load balancing is required when servers transfer data on the same path. This is necessary to avoid congestion. Load balancing is challenged by the dynamic transferral of demands and complex routing control. Because of the distributed nature of a traditional network, previous research on load balancing has mostly focused on improving the performance of the local network; thus, the load has not been optimally balanced across the entire network. In this paper, we propose a novel dynamic load-balancing algorithm for fat-tree. This algorithm avoids congestions to the great possible extent by searching for non-conflicting paths in a centralized way. We implement the algorithm in the popular software-defined networking architecture and evaluate the algorithm’s performance on the Mininet platform. The results show that our algorithm has higher bisection bandwidth than the traditional equal-cost multi-path load-balancing algorithm and thus more effectively avoids congestion.

关键词: data center network, software-defined networking, load balancing, network management