计算机工程与应用 ›› 2021, Vol. 57 ›› Issue (19): 116-122.DOI: 10.3778/j.issn.1002-8331.2009-0507

• 大数据与云计算 • 上一篇    下一篇

数据中心基于服务满足度的网络流避让方法

武自强,周建涛,赵大明,柳林   

  1. 1.内蒙古大学 计算机学院,呼和浩特 010021
    2.内蒙古云计算与服务软件工程实验室,呼和浩特 010021
    3.内蒙古数据处理与社会计算重点实验室,呼和浩特 010021
    4.内蒙古师范大学 计算机科学技术学院,呼和浩特 010021
  • 出版日期:2021-10-01 发布日期:2021-09-29

Network Flow Avoidance Method Based on Service Satisfaction in Data Center

WU Ziqiang, ZHOU Jiantao, ZHAO Daming, LIU Lin   

  1. 1.College of Computer Science, Inner Mongolia University, Hohhot 010021, China
    2.Inner Mongolia Engineering Laboratory for Cloud Computing and Service Software, Hohhot 010021, China
    3.Inner Mongolia Key Laboratory of Data Processing and Social Computing, Hohhot 010021, China
    4.College of Computer Science and Technology, Inner Mongolia Normal University, Hohhot 010021, China
  • Online:2021-10-01 Published:2021-09-29

摘要:

越来越多的行业开始利用云以降低成本提高生产力,支撑多样化的服务对数据中心的网络性能提出了更高的要求,如何高负载下优先保证各项服务的质量变得至关重要,云服务提供商同时也关注如何提高数据中心的网络资源利用率并降低能耗。结合上述问题,提出一种基于服务满足度对非服务网络流进行调度的方法。引入服务满足度这一概念,评估网络状态能否满足服务需要,然后依据网络流所支撑的不同服务将其分类,网络负载激增时基于服务满足度调整非用户服务依赖的网络流,降低网络负载缓解拥塞。仿真结果表明,主动避让方法在网络高负载时能优先保证服务的质量,同时提高网络性能。

关键词: 网络流, 服务满足度, 主动避让, 网络拥塞, 服务分类

Abstract:

More and more industries are starting to utilize the cloud to reduce costs and improve productivity. Supporting diversified service raises higher requirements on the network performance of data centers, which is critical to ensure the quality of various services under high load. Cloud service providers also pay attention to improving the utilization of network resources and reducing energy consumption of data centers. A method for scheduling non-service network flows based on service satisfaction is proposed to solve the problems mentioned above. Firstly, the concept of service satisfaction is introduced to assess whether the network status can meet the service needs, and then it is classified according to the different services supported by the network flow. When the network load surges, the network flow that is not dependent on user services will be adjusted based on service satisfaction to reduce network load and alleviate congestion. The simulation results show that the active avoidance method can give priority to ensuring the quality of service when the network is suffering heavy load and improve network performance at the same time.

Key words: network flow, service satisfaction, active avoidance, network congestion, service classification