On queue length moments in fork and join queuing networks with general service times

  • Simonetta Balsamo
  • Ivan Mura
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1245)


Fork and join queueing network models can be used to represent and analyse the performance of distributed and parallel processing computer systems with concurrence and synchronisation constraints and shared resources. Concurrence and synchronisation make the solution of such performance models more complex than classical queueing network analysis. We consider parallel processing systems where a set of independent tasks are executed simultaneously and can be represented by fork and join queueing networks. We present an approximate method to analyse open fork and join queueing networks with general service time distribution represented by Coxian distribution. Specifically, we derive two approximations of the joint queue length distribution and closed form expressions for all the moments of this distribution. We prove that the method provides lower and upper bounds on the queue length moments.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    F. Baccelli and A.M. Makowski, “Queueing models for systems with synchronization constraints” Proceedings of the IEEE, Vol. 77,1989.Google Scholar
  2. [2]
    F. Baccelli, W.A. Massey and D. Towsley, “Acyclic fork-join queueing network” J. of ACM, Vol. 36, 1989.Google Scholar
  3. [3]
    S. Balsamo and L. Donatiello, “Approximate performance analysis of parallel processing systems” in “Decentralised Systems”, M. Cosnard and C. Girault Ed., Elsevier Science, 1990, pp. 325–336.Google Scholar
  4. [4]
    S. Balsamo and I. Mura, “Approximate response time distribution in fork and join systems” in Proc. ACM Sigmetrics/Performance'95 Joint Conference, Ottawa, Ontario Canada, 1995.Google Scholar
  5. [5]
    S. Balsamo and I. Mura, “Synchronisation delay in hardware fault-tolerance techniques” in Proc. IEEE IPDS '96 Conf., Urbana-Champaign, Illinois USA, 1996.Google Scholar
  6. [6]
    S. Balsamo and I. Mura, “Bounds on queue length moments in Parallel Processing Systems with Coxian Services” University of Udine, Dept. of Math. and Comp. Sci., UDMI/04/97/RR, 1997.Google Scholar
  7. [7]
    M.A. Brun and G. Fajolle, “A distribution of the transaction processes in a simple fork and join system” in “Computer Performance and Reliability”, P. J. Courtois, G. Iazeolla and O. J. Boxma Ed., Elsevier Science, North-Holland, 1988.Google Scholar
  8. [8]
    D.R. Cox, “A use of complex probabilities in the theory of stochastic process” Proc. Cambridge Philos. Soc, Vol. 51, pp. 313–319, 1955.Google Scholar
  9. [9]
    L. Flatto and S. Hahn, “Two parallel queues created by arrivals with two demands I” SIAM Journal of Applied Mathematics, Vol. 44, pp. 1041–1053, 1984.CrossRefGoogle Scholar
  10. [10]
    C. Kim and A.K. Agrawala, “Analysis of the Fork-Join queue” IEEE Transactions on Computers, Vol. 38, pp. 250–255, 1989.CrossRefGoogle Scholar
  11. [11]
    D.E. Knuth, “The art of computer programming” Addison-Wesley, 1973.Google Scholar
  12. [12]
    R. Nelson, A. Tantawi and D. Towsley, “Performance analysis of parallel processing systems” IEEE Transactions of Software Engineering, Vol. 14, 1988.Google Scholar
  13. [13]
    R. Nelson and A.N. Tantawi, “Approximate analysis of fork-join synchronisations in parallel queues” IEEE Transactions on Computers, Vol. 37, pp. 739–743, 1988.CrossRefGoogle Scholar
  14. [14]
    M. F. Neuts, “Matrix-geometric solutions in stochastic models” Johns Hopkins University Press, 1981.Google Scholar
  15. [15]
    H.G. Perros, “On the M/Ck/l queue” Performance Evaluation, Vol. 3, pp. 83–93, 1983.CrossRefGoogle Scholar
  16. [16]
    B.M. Rao and M.J.M. Posner, “Algorithmic approximation analysis of the split and match queue” Comm. Statis. Stochastic Models, Vol. 1, pp. 433–456, 1985.Google Scholar
  17. [17]
    N. Van Dijk, “Truncation of Markov Chains with Applications to Queueing” Oper. Res., Vol. 39, 1991.Google Scholar
  18. [18]
    N. Van Dijk, “Error Bound Analysis of Queueing Networks” in Tutorial of Performance'96, Lausanne, Switzerland, 1996.Google Scholar
  19. [19]
    S. Varma and A.M. Makowski, “Interpolation approximation for symmetric fork-join queues” in Proc. Performance '93, Rome, 1993, pp. 245–273.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • Simonetta Balsamo
    • 1
  • Ivan Mura
    • 2
  1. 1.Dept. of Mathematics and Computer ScienceUdineItaly
  2. 2.Dept. Information EngineeringUniversity of PisaPisaItaly

Personalised recommendations