One of the main objectives in the design of a file system is the reduction of storage and data transfer costs. This paper presents a model in which several requests access the Ble system, and each request requires information from one or more variable length dataitems. The probabilities of access and the distribution of each data-item's length are assumed to be known, and to be mutually independent. The file system uses one or more storage devices, and each record may be partitioned into subrecords that are stored on different devices. One of the suhrecords is designated as the primary record; when a request for a record is made, the primary record is first accessed, and other subrecords are accessed only if the pertinent information is not stored in the primary record. The model that is presented in this paper, both as a nonlinear programming model and a mixed integer programming model, is a very general one; several types of file systems may be derived from it by an appropriate selection of its parameters. This model has already been used in the optimization of library routines' storage at a large scale operating system.
Computer systems play an increasingly important role in modern enterprises. Such systems represent a major investment, and their performance has major effects on profitability, utilization and user satisfaction. Much effort is therefore directed at the efficient management of these systems, and this paper is a step in this direction. Specifically, we consider the optimal allocation of programs and data to various storage devices in order to minimize the expected operational costs, subject to capacity and timing constraints. This problem is one of the many optimization problems that the system management faces, and its solution, though suboptimal for the whole system, might contribute considerably to the overall system's performance and costs. An effective procedure for the optimal (storage) allocation is presented, whose application to an actual large-scale operating system resulted in high (relative) savings. This procedure, with minor modifications, might be applied to many allocation problems which face users and management of computer systems.
Consider a multipriority batch computer system which users from several different classes may join, with toll, service, and waiting charges. Such a system is formulated here as a semi-Markov decision process, in which the aim of arriving users is to minimize their expected loss. The optimal joining policy of arriving users who may join the system at some of its queues is a control limit policy, with a single control number for any possible queue and the user's class; a newly arriving user will join a queue that is not filled up to the control number corresponding to this queue and the user's class. In this paper control numbers, as well as lower and upper bounds for the control numbers and the capacities of the system's queues, are derived.
A generalized multi-entrance and multipriority M/G/1 time-sharing system is dealt with. The system maintains many separate queues, each identified by two integers, the prmnty level and the entry level The arrival process of users is a homogenous Polsson process, while service requirements are identically distributed and have a finite second moment Upon arrival a user joins one of the levels, through the entry queue of this level In the (n, k)-th queue, where n is the priority level and k is the entry level, a user m eligible to a (finite or lnfimte) quantum of service. If the service requirements of the user are satmfied during the quantum, the user departs, and otherwise he is transferred to the end of the (n ~ 1, k)-th queue for addltmnal service When a quantum of service is completed, the highest priority nonempty level is chosen to be served next, within thin level the queues are scanned according to the prmmty of their entry level, and the user at the head of the highest prmnty nonempty queue is chosen to be served In such a priority dlsclphne, preferred users always get an improved service though the serwce of all users is degraded m proportion to their service requirements Expected flow times and expected number of waiting users are demved and then specialized to the head-of-the-hne M/G/1 prmrlty dlsciphne (in whmh quanta have infinite length and serwce is uninterrupted) and to the FB, time-sharing system. Finally, the generahzed multientrance and multipriomty time-sharing dlsclphne is (numermally) compared with several other time-sharing systems.KEY WORDS AND PHRASES. time-sharing, multi-entrance prmrity queue CR CATEGORIES' 4 32, 4 35 IntroductwnSeveral models of time-sharing computer systems have been described and analyzed in recent years. In these models jobs with short service requirements receive enhanced service at the expense of degraded service for longer requirements; the various models differ in the degree of enhancement they provide. In the simplest time-sharing system (Figure 1), known as a round-robin system, a newly arriving user joins the end of a single queue. This queue is served by a single service facility according to an FIFO discipline. The maximum interval of service time which is given to any user is of limited length, and is called a quantum; if the user's request is fulfilled during this quantum he leaves the system. Otherwise, he joins the end of the queue, behind any recycled users and newly arriving users since the most recent moment at which he entered the queue. Such a system was studied by Kleinrock [8], Adiri and Avi-Itzhak [3], Adiri [2]', and others (see McKinney [11] for additional references).The scheduling discipline of a round-robin system ignores the amount of service, measured in number of service quanta, already given to users already present in the sys-
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.