A decomposable model of program paging behaviour

SummaryExisting models of program behaviour are shown to give an incomplete account of program locality. A model based on the distinction between short- and long-run equilibrium states in nearly completely decomposable systems is proposed to overcome this deficiency. This distinction leads to the combined use of a Markovian model of the transitions between localities and of separate models for the locality short-term behaviours. This combination is shown to give better estimations of the page fault rate and of the working set size distribution. The conditions under which this distribution is approximately normal and under which the assumptions of independent page references are valid are also clarified. The approach is illustrated by a numerical example, showing in particular that other models presented in the literature may have computer time and space requirements which are beyond practical possibilities.

[1]  Irving L. Traiger,et al.  Evaluation Techniques for Storage Hierarchies , 1970, IBM Syst. J..

[2]  Alfred V. Aho,et al.  Principles of Optimal Page Replacement , 1971, J. ACM.

[3]  P.-J. Courtois.  On the Near-Complete-Decomposability of Networks of Queues and of Stochastic Models of Multiprogramming Computing Systems , 1971 .

[4]  Jaroslav Král.  One way of estimating frequencies of jumps in a program , 1968, CACM.

[5]  Erol Gelenbe,et al.  A Unified Approach to the Evaluation of a Class of Replacement Algorithms , 1973, IEEE Transactions on Computers.

[6]  W. Frank King,et al.  Analysis of Demand Paging Algorithms , 1971, IFIP Congress.

[7]  Mark A. Franklin,et al.  Computation of page fault probability from program transition diagram , 1974, CACM.

[8]  Peter J. Denning,et al.  A study of program locality and lifetime functions , 1975, SOSP.

[9]  Peter Naur.  The performance of a system for automatic segmentation of programs within an ALGOL compiler (GIER ALGOL) , 1965, CACM.

[10]  Peter J. Denning,et al.  Experiments with program locality , 1899, AFIPS '72 (Fall, part I).

[11]  Ulf Grenander,et al.  Patterns in Program References , 1975, IBM J. Res. Dev..

[12]  Peter Bryant,et al.  Predicting Working Set Sizes , 1975, IBM J. Res. Dev..

[13]  P. Courtois.  Error Analysis in Nearly-Completely Decomposable Stochastic Systems , 1975 .

[14]  A. Palmboom,et al.  Analysis of Demand Paging Algorithms , 1972, GI Jahrestagung.

[15]  Peter J. Denning,et al.  Virtual memory , 1970, CSUR.

[16]  Donald J. Hatfield,et al.  Program Restructuring for Virtual Memory , 1971, IBM Syst. J..

[17]  Peter J. Denning,et al.  The working set model for program behavior , 1968, CACM.

[18]  Herbert A. Simon,et al.  The Sciences of the Artificial , 1970 .

[19]  Gerald S. Shedler,et al.  Locality in Page Reference Strings , 1972, SIAM J. Comput..

[20]  Peter J. Denning,et al.  Properties of the working-set model , 1972, CACM.

[21]  Herbert A. Simon,et al.  Aggregation of Variables in Dynamic Systems , 1961 .

[22]  E. Gelenbe.  B73-10 Time Sharing Systems , 1973, IEEE Transactions on Computers.

[23]  Laszlo A. Belady,et al.  A Study of Replacement Algorithms for Virtual-Storage Computer , 1966, IBM Syst. J..

[24]  Alan P. Batson,et al.  Measurements of major locality phases in symbolic reference strings , 1976, SIGMETRICS '76.

[25]  Alan P. Batson,et al.  Characteristics of program localities , 1976, CACM.

[26]  Hendrik Vantilborgh.  On the working set size and its normal approximation , 1974 .

[27]  Juan Rodriguez-Rosell,et al.  Empirical working set behavior , 1973, Commun. ACM.