Advertisement

Stability Analysis of a MAP/M/s Cluster Model by Matrix-Analytic Method

  • Evsey Morozov
  • Alexander RumyantsevEmail author
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9951)

Abstract

In this paper, we study the stability conditions of the multiserver system in which each customer requires a random number of servers simultaneously and a random service time, identical at all occupied servers. We call it cluster model since it describes the dynamics of the modern multicore high performance clusters (HPC). Stability criterion of an M/M/s cluster model has been proved by the authors earlier. In this work we, again using the matrix-analytic approach, prove that the stability criterion of a more general MAP/M/s cluster model (with Markov Arrival Process) has the same form as for M/M/s system. We verify by simulation that this criterion (in an appropriate form) allows to delimit stability region of a MAP/PH/s cluster model with phase-type (PH) service time distribution. Finally, we discuss asymptotic results related to accelerated stability verification, as well as to the new method of accelerated regenerative estimation of the performance metrics.

Keywords

Stability condition High performance cluster Map arrivals Simultaneous service multiserver system 

Notes

Acknowledgments

This research is partially supported by Russian Foundation for Basic Research, grants 15-07-02341, 15-07-02354, 15-07-02360, 15-29-07974, 16-07-00622 and the Program of Strategic Development of Petrozavodsk State University. The authors thank Udo Krieger for a few useful comments.

References

  1. 1.
    Brill, P., Green, L.: Queues in which customers receive simultaneous service from a random number of servers: a system point approach. Manag. Sci. 30(1), 51–68 (1984)MathSciNetCrossRefzbMATHGoogle Scholar
  2. 2.
    Chakravarthy, S., Karatza, H.: Two-server parallel system with pure space sharing and Markovian arrivals. Comput. Oper. Res. 40(1), 510–519 (2013)MathSciNetCrossRefGoogle Scholar
  3. 3.
    Federgruen, A., Green, L.: An M/G/c queue in which the number of servers required is random. J. Appl. Probab. 21(3), 583–601 (1984)MathSciNetCrossRefzbMATHGoogle Scholar
  4. 4.
    Feitelson, D.G.: Workload Modeling for Computer Systems Performance Evaluation. Cambridge University Press, New York (2015)CrossRefzbMATHGoogle Scholar
  5. 5.
    Filippopoulos, D., Karatza, H.: An M/M/2 parallel system model with pure space sharing among rigid jobs. Math. Comput. Model. 45(5–6), 491–530 (2007)MathSciNetCrossRefzbMATHGoogle Scholar
  6. 6.
    Gillent, F., Latouche, G.: Semi-explicit solutions for M/PH/1-like queuing systems. Eur. J. Oper. Res. 13(2), 151–160 (1983)MathSciNetCrossRefzbMATHGoogle Scholar
  7. 7.
    He, Q.-M.: Fundamentals of Matrix-Analytic Methods. Springer, New York (2014)CrossRefzbMATHGoogle Scholar
  8. 8.
    Ibe, O.C.: Markov Processes for Stochastic Modeling. Academic Press, Amsterdam, Boston (2009)zbMATHGoogle Scholar
  9. 9.
    Kim, S.S.: M/M/s queueing system where customers demand multiple server use. Ph.D. thesis, Southern Methodist University (1979)Google Scholar
  10. 10.
    Latouche, G., Ramaswami, V.: Introduction to Matrix Analytic Methods in Stochastic Modeling. ASA-SIAM, Philadelphia (1999)CrossRefzbMATHGoogle Scholar
  11. 11.
    Morozov, E., Nekrasova, R., Peshkova, I., Rumyantsev, A.: A regeneration-based estimation of high performance multiserver systems. In: Gaj, P., Kwiecien, A., Stera, P. (eds.) CN 2016. CCIS, vol. 608, pp. 271–282. Springer, Heidelberg (2016). doi: 10.1007/978-3-319-39207-3_24 CrossRefGoogle Scholar
  12. 12.
    Neuts, M.F.: Matrix-Geometric Solutions in Stochastic Models. Johns Hopkins University Press, Baltimore (1981)zbMATHGoogle Scholar
  13. 13.
    Neuts, M.F., Dan, L., Surya, N.: Local poissonification of the Markovian arrival process. Commun. Stat. Stoch. Model. 8, 87–129 (1992)MathSciNetCrossRefzbMATHGoogle Scholar
  14. 14.
    R Foundation for Statistical Computing. Vienna, Austria. ISBN: 3-900051-07-0. http://www.r-project.org/
  15. 15.
    Rumyantsev, A.: hpcwld: High Performance Cluster Models Based on Kiefer-Wolfowitz Recursion. http://cran.r-project.org/web/packages/hpcwld/index.html
  16. 16.
    Rumyantsev, A., Morozov, E.: Accelerated verification of stability of simultaneous service multiserver systems. In: 2015 7th International Congress on Ultra Modern Telecommunications and Control Systems and Workshops (ICUMT), pp. 239–242. IEEE (2015)Google Scholar
  17. 17.
    Rumyantsev, A., Morozov, E.: Stability criterion of a multiserver model with simultaneous service. Ann. Oper. Res., 1–11 (2015). doi: 10.1007/s10479-015-1917-2
  18. 18.
    Rumyantsev, A.: An HPC upgrade/downgrade that provides workload stability. In: Malyshkin, V. (ed.) PaCT 2015. LNCS, vol. 9251, pp. 279–284. Springer, Heidelberg (2015)CrossRefGoogle Scholar
  19. 19.
    Scheller-Wolf, A., Vesilo, R.: Sink or swim together: necessary and sufficient conditions for finite moments of workload components in FIFO multiserver queues. Queueing Syst. 67(1), 47–61 (2011)MathSciNetCrossRefzbMATHGoogle Scholar
  20. 20.
    Sutter, H., Larus, J.: Software and the concurrency revolution. Queue 3(7), 54–62 (2005)CrossRefGoogle Scholar
  21. 21.
    Van Dijk, N.M.: Blocking of finite source inputs which require simultaneous servers with general think and holding times. Oper. Res. Lett. 8(1), 45–52 (1989)MathSciNetCrossRefzbMATHGoogle Scholar
  22. 22.
    Wagner, D., Naumov, V., Krieger, U.R.: Analysis of a multi-server delay-loss system with a general Markovian arrival process. Techn. Hochschule, FB 20, Inst. für Theoretische Informatik, Darmstadt (1994)Google Scholar

Copyright information

© Springer International Publishing AG 2016

Authors and Affiliations

  1. 1.Institute of Applied Mathematical ResearchKarelian Research Centre of RASPetrozavodskRussia
  2. 2.Petrozavodsk State UniversityPetrozavodskRussia

Personalised recommendations