S. S. Lavenberg. Computer Performance Modeling Handbook. Academic Press, New York, USA, 1983.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....for the second example above, i.e. 1 and = 10 with z in the x axis and the loss probability in the y axis. The eight curves correspond to c = 0:125; 0:25; 0:375; 0:5; 0:625; 0:75; 0:875 and 1.0. We use the regenerative method to generate the con dence interval for the simulations [21] and halt the simulation when the ratio of the width of the 99 con dence interval to the estimated loss probability is less than 0.005. A curve tting technique, based on an approximation of the distribution of the delay through the lter, is used to nd the optimal z s in these curves. Though ....

S.S. Lavenberg, "Computer Performance Modeling Handbook," Academic Press, 1983.

....similar to SP graphs. In order to express non SP structures Pamela also offers explicit condition synchronization operators [23] We conclude this section by an example that forms a typical demonstration of the modeling approach in Pamela . Example 3.1. 1 Consider a machine repair model (MRM) [30] in which P clients either spend l on local processing, or request service from a single FCFS server s with service time s , for a total cycle count of N iterations (unlike the steady state analysis typical for queuing systems or Petri nets, in our approach we require models to terminate) The ....

S.S. Lavenberg, Computer Performance Modeling Handbook. New York: Academic Press, 1983. ISBN 0-12-438720-9.

....time l doing local processing or request (RPC like) service from a central server s, with service time s , with a total visit count of N Practical examples are CPU, DMA, switch, memory, but also critical sections in programs. iterations. The Pamela model of this machine repair model (MRM) [23] is given by L = par (p = 1 : P ) seq (i = 1 : N) f delay( l ) use(s; s ) in which the exclusive service is expressed by the use operation applied to resource s representing the server. Note, that the above mathematical expression L is displayed in program format, including the ....

S.S. Lavenberg, Computer Performance Modeling Handbook. New York: Academic Press, 1983.

....cases. However, we believe that the modeling framework described in the previous section is certainly addressing the first one of the aforementioned limitations and it is also amenable to appropriately chosen extensions for addressing the other two. We suggest the use of a number of well known ([14]) extended queuing modeling constructs, for the representation of blocking, simultaneous resource possession and parallel execution flows phenomena. Their use is illustrated in the context of CORBA based object interaction cases and is supported by the modeling tools mentioned so far. More ....

S. S. Lavenberg (Ed.), Computer Performance Modeling Handbook, Academic Press, 1983.

....various kinds of dependencies that occur in reliability availability models [13, 14, 44] Task precedence graphs [21, 22, 26, 37, 38] can be used for the performance analysis of concurrent programs with unlimited system resources. System performance analysis using product form queueing networks [23, 25] can instead consider contention for resources. The product form assumptions are not satis ed, however, when behaviors such as concurrency within a job, synchronization, and server failures are considered. Once again, Markov models do provide a framework to address all these concerns. The common ....

S. S. Lavenberg, editor. Computer Performance Modeling Handbook. Academic Press, 1983.

....is no demand for service, and the thread pool shrinks. Rather than imposing a fixed limit on the number of threads in a stage, the thread pool controller can automatically determine the maximum thread pool size, using mechanisms similar to multiprogramming level control in virtual memory systems [82]. The controller samples the throughput of each stage (requests processed per second) maintaining a weighted moving average 103 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 160 180 0 100 200 300 400 500 600 700 800 900 Throughput, requests sec Number of threads Time ....

S. S. Lavenberg. Computer Performance Modeling Handbook. Academic Press, 1983.

....of these stereotypes are shown. For sake of conciseness and readability, we do not discuss the details of the dispatchers State Diagrams, rather we focus on the CPU one. The CPU is modeled as a queued service center that extracts jobs from the queue following a quantum based round robin strategy [7, 8]. In the idle state the queue is supposed to be empty and no job is being served. Upon the arrival of a job, the CPU becomes busy and it returns to the idle state in any moment the queue is idle and no job is being served. Two state transitions originate from the busy state. In case of a new ....

Lavenberg, S.S. "Computer Performance Modeling Handbook", Academic Press, New York, 1983.

....the Finite Time LP. It provides transient bounds on the performance of the system, from any initial condition. 1 Introduction Queueing networks are a useful class of models in many application domains, e.g. communication networks [1] manufacturing systems [2] 3] and computer systems [4]. Control is typically exercised over such systems by the use of scheduling policies; see [5] However, if one ventures outside a certain class of special systems for which the steady state distribution has a product form, very little is known concerning their performance, or even their ....

Stephen Lavenberg, Editor, Computer Performance Modeling Handbook. New York: Academic Press, 1983.

....also dene his own distribution. Seawind uses Lehmer s algorithm, with an additional shuing to generate uniformly distributed random numbers and output random values that conform to the chosen distribution. Lehmer s algorithm has been studied very intensively, for example in [PaMi88] EaCS94] and [Lave83]. The additional shuing is taken from [Knut98] pages 142 146. The implementations of the other random distributions are based on algorithms presented in [LaKe91] For all random distributions, the upper and lower limits can also be set, so that the wireless network delay never falls below or ....

Lavenberg, S. S., Computer Performance Modeling Handbook. Academic Press, New York, 1983, pp. 223-229.

....is well within tens of percents. Although applied in the sequel, due to space considerations the cost estimation algorithm itself is not described in the paper. Details can be found in [4] 2. 2 Contention Modeling A classical example in performance modeling is the machine repair model (MRM) [9] in which P clients either spend a mean time l on local processing, or request service from a server s (s = 1) with mean service time s , for a total cycle count of N iterations. Both time delays are typically stochastic. The Pamela model of the MRM is given by L = par (p = 1; P ) seq (i = 1; ....

S.S. Lavenberg, Computer Performance Modeling Handbook. New York: Academic Press, 1983. ISBN 0-12-438720-9.

....numerical transform inversion, scaling, dimension reduction, Euler summation. 1. Introduction For over twenty years, closed queueing networks and related product form models have played a major role in the performance analysis of computer systems, communication networks and other complex systems [5,14,19,28,38]. The success of these models has largely been due to the excellent algorithms for computing the steady state performance measures that have been developed, such as the convolution algorithm [6,35] the mean value analysis (MVA) algorithm [36] the tree convolution algorithm [27] the recursion by ....

.... such as the convolution algorithm [6,35] the mean value analysis (MVA) algorithm [36] the tree convolution algorithm [27] the recursion by chain algorithm (RECAL) 12,13] the mean value analysis by chain (MVAC) algorithm [15] and the distribution analysis by chain (DAC) algorithm [37] see [5,14,28] for an overview. While these algorithms for closed queueing networks have significant differences, they can all be related by their common recursive approach [14] These algorithms have been very successful, but they do encounter difficulties when the model becomes large in one way or another, ....

[Article contains additional citation context not shown here]

LAVENBERG, S. S. (ed.) Computer Performance Modeling Handbook, Academic Press, Orlando, FL, 1983.

....and the scheduler in terms of queueing networks. Queueing models have their origin in the study of design problems of automatic telephone exchanges. They are widely used to evaluate the performance of (amongst other) manufacturing systems, communication systems and computer systems. In Lavenberg [1], many examples of queueing models for computer systems can be found. We use stochastic analysis to approximate the mean, variance and also higher moments of the transaction response time (response time is the time from transaction request to transaction termination) This allows us to estimate ....

S.S. Lavenberg. Computer Performance Modeling Handbook. Academic Press, Orlando, Fla., 1983.

....of Lucent Technologies, 1600 Mass. Ave. #707, Cambridge MA, 02138. e mail: awberger lucent.com. AT T Labs, 200 Laurel Ave. Middletown, NJ 07748. e mail yakov buckaroo.mt.att.com. Berger Kogan 2 09 20 99 computer systems and networks, see for example [1] or [2] and our model is one of the simplest; though the entity modeled by a job is non standard, see Section 2. Our CQN model consists of one inifinite server (IS) and one processor sharing (PS) server and was motivated by Heyman, Lakshman, and Neidhardt, 3] who also use a CQN model where ....

S. Lavenberg, Ed. Computer Performance Modeling Handbook, Academic Press, 1983.

....let (r , i ) C 1 . We start with a standard expression for the probability mass function of q 1i , the number of chain 1 customers at queue i, namely, P(q 1i = k) g(K) r 1i k (g(K k1 1 ) r 1i g(K (k 1)1 1 ) 6.1) see (3.257) on p. 147 of [28]. A similar expression holds for the mass function of n ri [28] It involves r ri instead of r 1i . From the telescoping property of (6.1) we can write the tail probabilities as P(q 1i = k) g(K) r 1i k g(K k1 1 ) 6.2) From (6.2) we obtain the standard formula for the ....

....probability mass function of q 1i , the number of chain 1 customers at queue i, namely, P(q 1i = k) g(K) r 1i k (g(K k1 1 ) r 1i g(K (k 1)1 1 ) 6.1) see (3.257) on p. 147 of [28] A similar expression holds for the mass function of n ri [28]. It involves r ri instead of r 1i . From the telescoping property of (6.1) we can write the tail probabilities as P(q 1i = k) g(K) r 1i k g(K k1 1 ) 6.2) From (6.2) we obtain the standard formula for the mean, E[q 1i ] k = 1 S P(q 1i k) k = 1 S K 1 r ....

[Article contains additional citation context not shown here]

LAVENBERG, S. S. (ed.) Computer Performance Modeling Handbook, Academic Press, Orlando, FL, 1983.

.... equations, time dependent behaviour 1991 Mathematics Subject Classi cation: 60K25 1 corresponding author 1 Introduction Queueing network models have been extensively used to represent and analyze resource sharing systems such as production, communication and information systems ( 7] 8] [11]) Queueing networks have been analyzed and solved under di erent assumptions and constraints. Various classes of queueing networks have product form solutions and ecient algorithms to exactly evaluate their performance have been developed [6] Queueing networks with blocking are used to represent ....

Stephen S. Lavenberg (Ed.). Computer Performance Modeling Handbook. Academic press, New York (1983).

....Telektronikk 2 1995, Specail issue on teletraffic quantity of interest is known, and therefore the results of Section 5.4 should be very flattering and in favour of transition splitting. The basic idea is simply to assume that in every simulation replica (using regenerative simulation, see e.g. [24]) he starts with a sequence of upward transitions from empty until a full system is reached, and then simulate the sequence of events (transitions) from this point and back to an empty system again. The sequence of upward transitions until full system without returning to an empty system first, is ....

S. Lavenberg. Computer Performance Modeling Handbook. Academic Press, 1983.

No context found.

S. S. Lavenberg. Computer Performance Modeling Handbook. Academic Press, New York, USA, 1983.

No context found.

S. S. Lavenberg. Computer Performance Modeling Handbook. Academic Press, New York, USA, 1983.

No context found.

S. S. Lavenberg. Computer Performance Modeling Handbook. Academic Press, New York, USA, 1983.

No context found.

S. S. Lavenberg, ed., Computer Performance Modeling Handbook. Academic Press, 1983.

No context found.

S. S. Lavenberg, Ed., Computer Performance Modeling Handbook, Academic Press, 1983.

No context found.

S. S. Lavenberg, Computer Performance Modeling Handbook, Academic Press (1983).

No context found.

S. S. Lavenberg, Ed., Computer Performance Modeling Handbook, Academic Press, 1983.

No context found.

Stephen Lavenberg, Editor. Computer Performance Modeling Handbook. Academic Press, New York, 1983.

No context found.

LAVE83 Lavenberg S. (Ed.), Computer Performance Modeling Handbook, Academic Press, Orlando, Florida, 1983.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC