The Journal of Korean Institute of Information Technology (한국정보기술학회논문지)

Korean Institute of Information Technology (한국정보기술학회)
권호 표지

Performance Analysis of MapReduce Application on Private Cloud by using OpenStack

OpenStack을 활용한 사적 클라우드에서 MapReduce 응용 성능분석

  • Kang, Yun-Hee (Div. of Information and Communication, Basekseok University)
  • Received : 2013.09.14
  • Accepted : 2013.10.17
  • Published : 2013.12.31
⟨ Previous Next ⟩

Abstract

Recently software based virtualization technology has been used to utilize high performance hardware computing resources for providing high availability. MapReduce programming has been driving Internet services and those services operation in a private cloud environment, which is composed of virtualized computing resources. Hence it is required to efficiently provide virtualized resources for handling diverse MapReduce applications. In this paper OpenStack is used to build a private cloud. We show the relationship of a Hadoop application and virtualized resource allocations in the experiment. As a result of this experiment, the performance of Hadoop based compute intensive application provides up to 110.76% of the performance of physical server when running on the virtual machine in the private cloud.

최근 가상화 기술은 고성능 하드웨어 컴퓨팅 자원의 가용성을 높이기 위한 소프트웨어 기술로 사용되고 있다. MapReduce 프로그래밍은 가상화 컴퓨팅 자원으로 이루어진 사적 클라우드 컴퓨팅 환경에서 인터넷 서비스 및 서비스 운영을 주도하고 있다. 이에 다양한 MapReduce 응용에 필요한 가상화된 자원의 효율적인 제공이 요구된다. 이를 위해 본 논문에서는 OpenStack을 사용하여 사적 클라우드를 구성한 후 MapReduce 기반 Hadoop 응용의 성능 분석을 수행하여 가상자원 할당과의 상관관계를 보인다. 실험결과 분석을 통해 Hadoop 기반 계산 중심 응용은 사적 클라우드 내의 가상 머신에서 수행한 경우 물리서버 수행 대비 최대 110.76%의 성능비를 얻을 수 있었다.

Keywords

References

  1. M. Armbrust, A. Fox, and R. Griggith, et al., "Above the cloud: A Berkeley View of Cloud Computing", Technical Report No.UCB/EECS- 2009-28, EECS Department, University of California at Berkeley, USA, Feb. 2009.
  2. Shadi Ibrahim, Hai Jin, Lu Lu, Li Qi, Song Wu, and Xuanhua Shi, "Evaluating MapReduce on Virtual Machines: The Hadoop Case", In Proceedings of the 1st International Conference on Cloud Computing (CloudCom '09), pp. 519-528, 2009.
  3. J. Dean and S. Ghemawat, "MapReduce: A Flexible Data Processing Tool", Communications of the ACM, Vol. 53, pp. 72-77, Jan. 2010.
  4. Yun-Hee Kang, "Construction of a MapReduce Application Running on Twister in Cloud Computing Environments, FutureGrid", Journal of KIIT, Vol. 9 No. 4, pp. 147-154, April 2011.
  5. J. Ekanayake, et al., "MapReduce for Data Intensive Scientific Analyses", the 2008 Fourth IEEE International Conference on eScience, pp. 277-284, Dec. 2008.
  6. M. Mahjoub and A. Mdhaffar, et al. "A Comparative Study of the Current Cloud Computing Technologies and Offers", IEEE First Symposium on Network Cloud Computing and Applications, pp. 131-134, Nov. 2011.
  7. P. Sempolinski and D. Thain, "A Comparison and Critique of Eucalyptus, OpenNebula and Nimbus", 2nd IEEE International Conference on Cloud Computing Technology and Science, pp. 417-426, Nov. 2010.
  8. Hadoop. http://hadoop.apache.org/
  9. OpenStack. http://www.openstack.org/
  10. KVM. http://www.linux-kvm.org/
  11. RackSpace. http://www.rackspace.com/