Earticle

초록

영어

A new task scheduling algorithm based on Hadoop is proposed to optimize scheduling of resources problems under the Distributed cloud-computing platforms. The core idea of the algorithm is full reference to the current network conditions and treat it as an important reference for system task scheduling, with the bandwidth management ability ,SDN architecture allows us to allocate bandwidth according to a time slot strategy, then according to the operation completed sooner or later the time to decide whether the task assigned to the local node or low load of non-local node .In this way, we will not only ensure the locality of the task from a global perspective , but also assign tasks in an optimal way to efficiently. In the end, we do experiments based on the scheduler to verify the quality of scheduling algorithm.

목차

Abstract
 1. Introduction
 2. Hadoop Task Scheduling Analysis
 3. SHSA Task Scheduling Algorithm
  3.1. Model of Job Completion Time
  3.2. SHSA Task Scheduling Process
  3.3. The Design of the SHSA Task Scheduler
 4. Experiment and Analysis of Results
  4.1. Experimental Environment and Data
  4.2. Contrast and Analysis of Job Completion Time
  4.3. Contrast and Analysis of Local Data Rate
 5. Conclusion
 References

키워드

저자

• Zhe Li [ College of Technology, Hubei Engineering University, Xiao Gan, China / Department of Automatic Control, Northwestern Polytechnical University, Xi’an, China ]
• Zhi-Long Deng [ Department of Electronics and Information, Northwestern Polytechnical University, Xi’an, China ]
• Tian-Fan Zhang [ College of Technology, Hubei Engineering University, Xiao Gan, China / Department of Automatic Control, Northwestern Polytechnical University, Xi’an, China ]

참고문헌

발행기관

간행물

표지보기 관심저널 등록

• 간행물명

  International Journal of Grid and Distributed Computing

• 간기

  격월간

• pISSN

  2005-4262

• 수록기간

  2008~2016

• 십진분류

  KDC 505 DDC 605