Journal of the Korean Institute of Intelligent Systems (한국지능시스템학회논문지)

Korean Institute of Intelligent Systems (한국지능시스템학회)
권호 표지

Fuzzy-based Processor Allocation Strategy for Multiprogrammed Shared-Memory Multiprocessors

다중프로그래밍 공유메모리 다중프로세서 시스템을 위한 퍼지 기반 프로세서 할당 기법

  • 김진일 (한남대학교 정보통신 멀티미디어공학부) ;
  • 이상구 (한남대학교 정보통신 멀티미디어공학부)
  • Published : 2000.10.01
Download PDF
⟨ Previous Next ⟩

Abstract

In the shared-memory mutiprocessor systems, shared processing techniques such as time-sharing, space¬sharing, and gang-scheduling are used to improve the overall system utilization for the parallel operations. Recently, LLPC(Loop-Level Process Control) allocation technique was proposed. It dynamically adjusts the needed number of processors for the execution of the parallel code portions based on the current system load in the given job. This method allocates as many available processors as possible, and does not save any processors for the parallel sections of other later-arriving applications. To solve this problem, in this paper, we propose a new processor allocation technique called FPA(Fuzzy Processor Allocation) that dynamically adjusts the number of processors by fuzzifYing the amounts ofueeded number of processors, loads, and estimated execution times of job. The proposed method provides the maximum possibility of the parallism of each job without system overload. We compare the performances of our approaches with the conventional results. The experiments show that the proposed method provides a better performance.

공유메모리 다중프로세서 시스템은 전체적인 시스템 이용률을 높이기 위하여 병렬 작업시 시분할(time-sharing), 공간분할(space-sharing), 갱스케줄링과 같은 프로세서 자원 공유 기법을 사용한다. 최근에는 주어진 작업의 병렬 코드 부분의 실행을 위해서 시스템 작업부하를 기준으로 프로세서의 수를 동적으로 조절하는 루프단계 프로세스 제어(LLPC)할당 기법이 제안되었다. 이 기법은 작업에 가능한 많은 프로세서를 할당하기 때문에, 나중에 도착하는 작업의 병렬부분을 수행해야 할 프로세서를 남겨 두지 않는다. 이러한 문제를 해결하기 위해, 본 논문에서는 작업부하량, 작업수행예상시간, 프로세스의 수를 퍼지화하여 시스템의 부하량에 따른 퍼지규칙으로 새로운 프로세서 할당 기법인 FPA(Fuzzy-based Processor Allocation)를 제안한다. 또한, 시스템의 과부하 없이 각 작업에 대한 최대한의 병렬 가능성을 제공함으로써 기존의 할당 기법에 비해 우수한 성능을 보인다.

Keywords

References

  1. ACM Trans. Computer Systems v.10 no.1 Scheduler Activations: Effective Kernel Support for User-Level Management of Parallelism T. Anderson;B. Bershad;E. Lazowska;H. Levy
  2. IEEE Trans. on Computer v.45 no.10 An Architecture for Tolerating Processor Failures in Shared-Memory Multiprocessors Michel Banatre;Alain Gefflaut(etc)
  3. Proc. IPPS '95 workshop Job Scheduling Strategies for Parallel Processing A Salable Multi-Discipline, Multiple-Processor Scheduling Framework for IRIX J. Barton;N. Bitar
  4. Convex Architecture Reference Manual(C-series)
  5. J. Parallel and Distributed Computing v.16 Gang Scheduling Performance Benefits for Fine-Grain Synchronization D. Feitelson;L. Rudolph
  6. Proc. Conf. Measurement and Modeling of Computer Systems v.19 The Impact of Operating System Scheduling Policies and Synchronization Methods on the performance of Parallel Applications A. Gupta;A. Tucker;S. Urushibara
  7. IEEE Trans. on Fuzzy Systems v.5 no.5 An Algorithmic Approach for Fuzzy Inference C.J. Kim
  8. Computer v.27 no.2 Exploiting the Parallel Available in Loops D. Lilja
  9. Supercomputing Performance Analysis of Job Scheduling Policies in Parallel Supercomputing Environments V. Naik;S. Setia;M. Squillante
  10. Proc. 1994 Int. Conf. Parallel Processing v.Ⅱ Impact of Loop Granularity and Self-Preemption on the Performance of Loop Parallel Applications on a Multiprogrammed Shared-Memory Multiprocessor C. Natarajan;S. Sharma;R. Iyer
  11. Proc. of the 10th International Parallel Processing Symposium Maximizing Speedup Through Self-Tuning of Processor Allocation Thu D. Nguyen;Raj Vaswani;John Zahorijan
  12. Proc. Distributed Computing Systems Conf. Scheduling Techniques for Concurrent Systems J. Ousterhout
  13. Proc. Int. Conf. Parallel Processing Symp. Comparing Gang-Scheduling with Dynamic Space Sharing on Symmetric Multiprocessors Using Automatic Self-Allocating Threads(ASAT) C. Severance;R. Enbody
  14. Proc. Int. Conf. Parallel Processing v.1 Automatic Self-Allocating Threads on the Convex Exemplar C. Severance;R. Enbody;S. Wallach;B. Funkhouser
  15. Efficient Scheduling on Multiprogrammed Shared-memory Multiprocessors A. Tucker
  16. Job Scheduling Strategies for Parallel Processing v.949 Loop-Level Process Control : An Effective Processor Systems K. Yue;H. Lilja;D. Feitelson(ed.);L. Rudolph(ed.)
  17. IEEE Trans. v.8 no.12 An Effective Processor Allocation Strategy for Multiprogrammed Shared-Memory Multiprocessors K. K. Yue;D. J. Lilja
  18. Proc. int. Seminar Performance Distributed and Parallel Systems Spinning Versus Blocking in Parallel Systems with Uncertainty J. Zahorian;E. Lazowska;D. Eager