A class-constrained resource allocation problem is considered. In this problem, a set of M heterogeneous resources is to be allocated optimally among a set of L users belonging to K user classes. A set of class alloca...
ISBN:
(纸本)9780897912549
A class-constrained resource allocation problem is considered. In this problem, a set of M heterogeneous resources is to be allocated optimally among a set of L users belonging to K user classes. A set of class allocation constraints, which limit the number of users of a given class that could be allocated to a given resource, is imposed. An algorithm with worst case time complexity O(M (LM + M2 + LK)) is presented along with a proof of its correctness. This problem arises in many areas of resource management in computersystems, such as load balancing in distributed systems, transaction processing in distributed database systems, and session allocation in time-shared computersystems. We illustrate the behavior of this algorithm with an example where file servers are to be allocated to workstations of multiple classes.
This conferenceproceedings contains 28 papers, 4 of them are in extended abstract form. The main subjects are database systems performance measures, packet routing algorithms for integrated switching networks, parall...
详细信息
ISBN:
(纸本)089791225X
This conferenceproceedings contains 28 papers, 4 of them are in extended abstract form. The main subjects are database systems performance measures, packet routing algorithms for integrated switching networks, parallel discrete event simulation, multi-disk management algorithms, performance analysis of parallel processing systems shuffle-exchange interconnection networks, multistage interconnection networks, performance through memory, statistical approach to computer performance modeling, performance evaluation of LAN sorting algorithms, and delay analysis of a window tree conflict resolution algorithms in local area network environment. All papers are separately indexed and abstracted.
Models of discrete systems are often utilized to assist in computer engineering and procurement. The tools for modeling have been traditionally developed using either analytic methods or discrete event simulation. The...
ISBN:
(纸本)9780897912259
Models of discrete systems are often utilized to assist in computer engineering and procurement. The tools for modeling have been traditionally developed using either analytic methods or discrete event simulation. The research presented here explores the use of statistical techniques to augment and assist this basic set of tools.
Our work deals with the analysis of the queueing delays of buffered multistage Banyan networks of multiprocessors. We provide tight upper bounds on the mean delays of the second stage and beyond, in the case of infini...
ISBN:
(纸本)9780897912259
Our work deals with the analysis of the queueing delays of buffered multistage Banyan networks of multiprocessors. We provide tight upper bounds on the mean delays of the second stage and beyond, in the case of infinite buffers. Our results are validated by simulations performed on a network simulator constructed by us. The analytic work for network stages beyond the first, provides a partial answer to open problems posed by previous research.
Queueing network models are effective tools for capacity planning of computersystems. The base of all performance oriented questions is the characterization of the computer system workload. At the capacity planning l...
ISBN:
(纸本)9780897912259
Queueing network models are effective tools for capacity planning of computersystems. The base of all performance oriented questions is the characterization of the computer system workload. At the capacity planning level the workload is described in user-oriented terms. At the system level the queueing network model requires input parameters, which differ from the workload description at the capacity planning level. In this paper a general procedure to transform the parameters between these two levels is presented and applied to a case study. The effect on system performance of an increase in the use of an existing application system is analysed.
We investigate two schemes for placing data on multiple disks. We show that declustering (spreading each file across several disks) is inherently better than clustering (placing each file on a single disk) due to a nu...
ISBN:
(纸本)9780897912259
We investigate two schemes for placing data on multiple disks. We show that declustering (spreading each file across several disks) is inherently better than clustering (placing each file on a single disk) due to a number of reasons including parallelism and uniform load on all disks.
Two of the most important parameters of a computer are its processor speed and physical memory size. We study the relationship between these two parameters by experimentally evaluating the intrinsic memory and process...
ISBN:
(纸本)9780897912259
Two of the most important parameters of a computer are its processor speed and physical memory size. We study the relationship between these two parameters by experimentally evaluating the intrinsic memory and processor requirements of various applications. We also explore how hardware prices are changing the cost effectiveness of these two resources. Our results indicate that several important applications are “memory-bound,” i.e., can benefit more from increased memory than from a faster processor.
Random-access systems (RAS) for collision-type channels have been studied extensively under the assumption of an infinite population which generates a Poisson arrival process. If the population is finite and if the (p...
ISBN:
(纸本)9780897912259
Random-access systems (RAS) for collision-type channels have been studied extensively under the assumption of an infinite population which generates a Poisson arrival process. If the population is finite and if the (practically desirable) free-access channel-access protocol is used, then it is shown that the specification of a channel-exit protocol is crucial for the stability and the fairness of the RAS. Free-exit and blocked-exit protocols are analyzed and it is concluded that the p-persistent blocked-exit protocol provides the mechanisms to assure stability and fairness for a wide range of arrival process models.
The same application implemented on different systems will necessarily present different workloads to the systems. Characterizations of workloads intended to represent the same application, but input to models of diff...
ISBN:
(纸本)9780897912259
The same application implemented on different systems will necessarily present different workloads to the systems. Characterizations of workloads intended to represent the same application, but input to models of different systems, must also differ in analogous ways. We present a hierarchical method for characterizing a workload at increasing levels of detail such that every characterization at a lower level still accurately represents the workload at higher *** discuss our experience in using the method to feed the same application through a workload characterization “pipeline” to two different models of two different systems, a conventional relational database system and a logic-based distributed database system. We have developed programs that partially automate the characterization changes that are required when the system to be modeled changes.
暂无评论