Two-Level Replacement Decisions in Paging Stores

Abstract

Abstract-One of the primary motivations for implementing virtual memory is its ability to automatically manage a hierarchy of storage systems with different characteristics. The composite system behaves as if it were a single-level system having the more desirable characteristics of each of its constituent levels. In this paper we extend the virtual memory concept to within the top level of a two-level hierarchy. Here, the top level is thought of as containing two additional levels within it. This hierarchy is not a physical one, but rather an artificial one arising from the employment of two different replacement algorithms. Given two replacement algorithms, one of which has good performance but high implementation cost and the other poor performance but low implementation cost, we propose and analyze schemes that result in an overall algorithm having the performance characteristics of the former and the cost characteristics of the latter. We discuss the suitability of such schemes in the management of storage hierarchies that lack page reference bits.

Transient and Steady-state Regime of a Family of List-based Cache Replacement Algorithms

“…Explicit results under the IRM model for the cache content distribution and miss probability have been derived for FIFO and LRU [22,2], RANDOM [15] and some simple hybrids thereof [1,3]. LRU is shown to outperform FIFO in [29], while FIFO and RANDOM perform identical [15].…”

“…, h, are large, this Markov chain is well approximated by a continuous time Markov chain. If all items 3 It should be clear that this ODE is Lipschitz-continuous and hence has a unique solution. Moreover, with the initial conditions indicated, the solution of the ODE satisfy that, for all time t: n k=1 x k,i (t) = mi and h i=0 x k,i (t) = 1.…”

Complete or fast reference trace collection for simulating multiprogrammed workloads

“…It removes the access permissions from all but a few of the more recently used pages, thus ignoring the vast majority of references to those pages while capturing references to all other pages. Therefore, it trades accuracy for speed by ignoring references that are unlikely to be relevant to a main memory simulation [2,4]. We show, in Section 3, that kVMTrace reduces total overhead dramatically, bringing it to a 3-to 20-fold slowdown.…”

“…For larger caches such as main memories, the vast majority of references are to recently used pages and have little effect on replacement [2,4]. kVMTrace relies on this property to capture fewer references.…”

A time invariant working set model for independent reference

“…A number of papers [1,3,5,17,19,24] have derived expressions for the miss ratio for different memory replacement algorithms under the IRM model. Fagin [14] analyzed the effect of page size on the miss ratio.…”