A Rigorous Analysis for Set-Up Time Models – A Metric Perspective

  • Eitan Bachmat
  • Tao Kai Lam
  • Avner Magen
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4112)


We consider model based estimates for set-up time. The general setting we are interested in is the following: given a disk and a sequence of read/write requests to certain locations, we would like to know the total time of transitions (set-up time) when these requests are served in an orderly fashion. The problem becomes nontrivial when we have, as is typically the case, only the counts of requests to each location rather then the whole input, in which case we can only hope to estimate the required time. Models that estimate the set-up time have been suggested and heavily used as far back as the sixties. However, not much theory exists to enable a qualitative understanding of such models. To this end we introduce several properties through which we can study different models such as (i) super-additivity which means that the set-up time estimate decreases as the input data is refined (ii) monotonicity which means that more activity produces more set-up time, and (iii) an approximation guarantee for the estimate with respect to the worst possible time.

We provide criteria for super-additivity and monotonicity to hold for popular models such as the independent reference model (IRM). The criteria show that the estimate produced by these models will be monotone for any reasonable system. We also show that the IRM based estimate functions, upto a factor of 2, as a worst case estimate to the actual set-up time.

To establish our theoretical results we use the theory of finite metric spaces, and en route show a result of independent interest in that theory, which is a strengthening of a theorem of Kelly [4] about the properties of metrics that are formed by concave functions on the line.


Triangle Inequality Disk Drive Activity Vector Negative Type Renewal Model 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Aven, O.I., Coffman, E.G., Kogan, Y.A.: Stochastic analysis of computer storage. D. Reidel publishing, Dordrecht (1987)MATHGoogle Scholar
  2. 2.
    Coffman, E.G., Denning, P.J.: Operating systems theory. Prentice Hall, Englewood Cliffs (1973)Google Scholar
  3. 3.
    Grossman, D.D., Silverman, H.F.: Placement of records on a secondary storage device to minimize access time. J. of the ACM 20, 429–438 (1973)MATHCrossRefGoogle Scholar
  4. 4.
    Kelly, J.B.: Hypermetric spaces and metric transforms. In: Shisha, O. (ed.) Inequalities III, pp. 149–158 (1972)Google Scholar
  5. 5.
    Lovasz, L., Pyber, L., Welsh, D.J.A., Ziegler, G.M.: Combinatorics in pure mathematics. In: Handbook of combinatorics, pp. 2039–2082. MIT press, North Holland (1995)Google Scholar
  6. 6.
    Opderbeck, H., Chu, W.W.: The renewal model for program behavior. SIAM J. of computing 4, 356–374 (1975)MATHCrossRefMathSciNetGoogle Scholar
  7. 7.
    Vanichpun, S., Makowski, A.: The output of a cache under the independent reference model - where did all the locality of reference go? In: Proceedings of SIGMETRICS (2004)Google Scholar
  8. 8.
    Wong, C.K.: Algorithmic Studies in mass storage systems. Computer science press, Rockville (1983)Google Scholar
  9. 9.
    Yue, P.C., Wong, C.K.: On the optimality of the probability ranking scheme in storage applications. J. of the ACM 20, 624–633 (1973)MATHCrossRefMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Eitan Bachmat
    • 1
  • Tao Kai Lam
    • 2
  • Avner Magen
    • 3
  1. 1.Dept. of Computer ScienceBen Gurion University 
  2. 2.EMC cooperationHopkinton
  3. 3.Dept. of Computer ScienceUniversity of Toronto 

Personalised recommendations