Adaptive Dynamic Load Balancing Strategies for Network-based Cluster Systems

네트워크 기반 클러스터 시스템을 위한 적응형 동적 부하균등 방법

  • 정훈진 (일진방사선엔지니어링(주) 연구원) ;
  • 정진하 (인하대학교 전자공학과) ;
  • 최상방 (인하대학교 전자전기컴퓨터공학부)
  • Published : 2001.12.01

Abstract

Cluster system provides attractive scalability in terms of compution power and memory size. With the advances in high speed computer network technology, cluster systems are becoming increasingly competitive compared to expensive MPPs (massively parallel processors). Load balancing is very important issue since an inappropriate scheduling of tasks cannot exploit the true potential of the system and can offset the gain from parallelization. In parallel processing program, it is difficult to predict the load of each task before running the program. Furthermore, tasks are interdependent each other in many ways. The dynamic load balancing algorithm, which evaluates each processor's load in runtime, partitions each task into the appropriate granularity and assigns them to processors in proportion to their performance in cluster systems. However, if the communication cost between processing nodes is expensive, it is not efficient for all nodes to attend load balancing process. In this paper, we restrict a processor that attend load balancing by the communication cost and the deviation of its load from the average. We simulate various models of the cluster system with parameters such as communication cost, node number, and range of workload value to compare existing load balancing methods with the proposed dynamic algorithms.

클러스터 시스템을 계산능력과 메모리 크기에 있어서 바람직한 확장성을 제공한다. 또한 고속의 컴퓨터 네트워크 기술의 발달로 인해 클러스터 시스템은 값비싼 MPPs (Massively Parallel Processors)와 비교하여 경쟁력을 얻고 있다. 부적당한 작업 스케줄링은 시스템의 기능을 충분히 이용할수 없고, 병렬처리 프로그램에서 프로그램 실행 전에 각 태스크의 부하를 예측하기 어려우며, 태스크들은 다양한 형태로 서로 의존적이다. 동적 부하균등 알고리즘에서는 실행시간에 각 프로세서의 부하를 평가한 후, 각 태스크를 적절한 크기로 분할하고 그것들을 각 프로세서의 수행능력에 비례하여 클러스터 시스템에 할당한다. 그러나, 프로세싱 노드간의 통신비용이 높으면, 모든 노드들이 부하분산에 참여하는 것은 효율적이지 못하다. 본 논문에서는 부하분산에 참여하는 프로세서를 통신비용과 평균 부하로부터의 편차를 고려하여 제한하였다. 기존의 부하균등 방식과 제안된 동적 알고리즘을 비교하기 위하여, 통신비용, 노드 수, 그리고 부하의 범위와 같은 파라메터를 사용하여 다양한 모델의 클러스터 시스템에 관한 시뮬레이션을 수행하였다.

Keywords

References

  1. R. Buyya, High performance cluster computing: Architectures and systems, volume 1, Prentice Hall, 1999
  2. M. Cermele, M. Colajanni, and G. Necci, 'Dynamic load balancing of distributed SPMD computations with explicit message-passing,' Sixth Heterogeneous Computing Workshop, pp. 2-16, Apr. 1997 https://doi.org/10.1109/HCW.1997.581406
  3. M.H. Willebeek-LeMair and A.P. Reeves, 'Strategies for dynamic load balancing on highly parallel computers,' IEEE Transactions on Parallel and Distributed Systems, vol. 4, no. 9, pp. 979-993, Sep, 1993 https://doi.org/10.1109/71.243526
  4. D.E. Culler and J.P. Singh, Parallel computer architecture: A hardware/software approach, Morgan Kaufmann Publishers, 1999
  5. M.J. Zaki, W. Li, and S. Parthasarathy. 'Customized dynamic load balancing for a network of workstations,' Journal of Parallel and Distributed Computing, vol, 43, no. 2, pp. 156-162, June 1997 https://doi.org/10.1006/jpdc.1997.1339
  6. N. Nedeljkovic and M. Quinn, 'Data-parallel programming on a heterogeneous workstations,' First International Symposium on High-Performance Distributed Computing, pp. 28-36, Sep, 1992
  7. E.P. Markatos and T,J. LeBlanc. 'Using processor affinity in loop scheduling on shared-memory multiprocessors,' IEEE Trans. Parallel and Distributed Systems, vol. 4. no. 5, pp. 379-400, Apr. 1994 https://doi.org/10.1109/71.273046
  8. J. Arabe et al., 'Dome: Parallel programming in a heterogeneous multi-user environment,' Technical Report 95-137, Carnegie Mellon University, Apr. 1995
  9. W.Y. Lee, S.J. Hong, J. Kim. and S. Lee, 'A dynamic load balancing algorithm on switch-based networks,' 13th International Conference on Parallel and Distributed Computing Systems, Aug. 2000
  10. M. Cierniak, M.J Zaki, and W. Li, 'Compile time scheduling algorithms for a heterogeneous NOW,' The Computer Journal, vol. 40, no. 6, pp, 356-372, Dec. 1997 https://doi.org/10.1093/comjnl/40.6.356
  11. W. Cai, B.S. Lee, A. Heng, and L. Zhu 'A simulation study of dynamic load balancing for network-based parallel processing,' Third International Symposium on Parallel Architecture, Algorithms, and Networks, pp. 383-389, Dec. 1997 https://doi.org/10.1109/ISPAN.1997.645124