A Stochastic Model of Fragmentation in Dynamic Storage Allocation

Abstract

We study a model of dynamic storage allocation in which requests for single units of memory arrive in a Poisson stream at rate λ and are accommodated by the first available location found in a linear scan of memory. Immediately after this first-fit assignment, an occupied location commences an exponential delay with rate parameter μ, after which the location again becomes available. The set of occupied locations (identified by their numbers) at time t forms a random subset S t of {1,2, . . .}. The extent of the fragmentation in S t , i.e. the alternating holes and occupied regions of memory, is measured by (S t ) -|S t |. In equilibrium, the number of occupied locations, |S|, is known to be Poisson distributed with mean ρ = λ/μ. We obtain an explicit formula for the stationary distribution of max (S), the last occupied location, and by independent arguments we show that (E max (S) -E|S|)/E|S| → 0 as the traffic intensity ρ → ∞. Moreover, we verify numerically that for any ρ the expected number of wasted locations in equilibrium is never more than 1/3 the expected number of occupied locations.Our model applies to studies of fragmentation in paged computer systems, and to containerization problems in industrial storage applications. Finally, our model can be regarded as a simple concrete model of interacting particles [Adv. Math., 5 (1970) Abstract. We study a model of dynamic storage allocation in which requests for single units of memory arrive in a Poisson stream at rate A and are accommodated by the first available location found in a linear scan of memory. Immediately after this first-fit assignment, an occupied location commences an exponential delay with rate parameter/z, after which the location again becomes available. The set of occupied locations (identified by their numbers) at time forms a random subset S, of 1, 2, .}. The extent of the fragmentation in S,, i.e. the alternating holes and occupied regions of memory, is measured by max (St) -IStl. In equilibrium, the number of occupied locations, ISI, is known to be Poisson distributed with mean p A//x. We obtain an explicit formula for the stationary distribution of max (S), the last occupied location, and by independent arguments we show that (E max (S) EISI)/EISI-0 as the traffic intensity /9 . Moreover, we verify numerically that for any/9 the expected number of wasted locations in equilibrium is never more than 1 / 2 the expected number of occupied locations.Our model applies to studies of fragmentation in paged computer systems, and to containerization problems in industrial storage applications. Finally, our model can be regarded as a simple concrete model of interacting particles [Adv. Math., 5(1970), pp. 246-290].

Transient Analysis for State-Dependent Queues with Catastrophes

“…Further, the results concerning dynamic storage allocation in computers can be found in Coffman et al [20].…”

A multicriteria approach for material yard planning

“…The dynamic storage allocation problem has been proved by Garey and Johnson (1979) to be NP-complete. Coffman et al (1985) study a model in which requests for single units of memory arrive according to a Poisson process. Kierstead (1991) uses an on-line algorithm for colouring interval graphs to construct a polynomial time approximation algorithm.…”

Some Exact and Asymptotic Solutions to Single Server Models of Dynamic Storage

“…This is in contrast to the present model(s) in which the system's service capacity is constant, independent of the number . The infinite server storage model has been studied by many authors, including Kosten [15] and Coffman, Kadota and Shepp [7] ; it is also the subject of an entire monograph by Newell [17] , who uses classical probabilistic methods to analyze the limiting case → ∞. The approach taken in Ref.…”

“…The approach taken in Ref. [7] consists of breaking up the cells into two subsets, those ranked 1, 2, , m, and those ranked m + 1, m + 2, . Then if 1 ( 2 ) is the number of items stored among the first (second) set of cells (thus From the distribution of ( 1 , 2 ) it is possible, by varying m, to also obtain expressions for the distributions of W and of max S , but these tend to also be very complicated.…”