Time Domain Measurement Representation in Computer System Diagnostics and Performance Analysis

Performance and Reliability Analysis of Computer Systems: An Example-Based Approach Using the SHARPE Software Package provides a variety of probabilistic, discrete-state models used to assess the reliability and performance of computer and communication systems. The models included are combinatorial reliability models (reliability block diagrams, fault trees and reliability graphs), directed, acyclic task precedence graphs, Markov and semi-Markov models (including Markov reward models), product-form queueing networks and generalized stochastic Petri nets. A practical approach to system modeling is followed; all of the examples described are solved and analyzed using the SHARPE tool. In structuring the book, the authors have been careful to provide the reader with a methodological approach to analytical modeling techniques. These techniques are not seen as alternatives but rather as an integral part of a single process of assessment which, by hierarchically combining results from different kinds of models, makes it possible to use state-space methods for those parts of a system that require them and non-state-space methods for the more well-behaved parts of the system. The SHARPE (Symbolic Hierarchical Automated Reliability and Performance Evaluator) package is the `toolchest' that allows the authors to specify stochastic models easily and solve them quickly, adopting model hierarchies and very efficient solution techniques. All the models described in the book are specified and solved using the SHARPE language; its syntax is described and the source code of almost all the examples discussed is provided. Audience: Suitable for use in advanced level courses covering reliability and performance of computer and communications systems and by researchers and practicing engineers whose work involves modeling of system performance and reliability.

Computation of bounds for transient measures of large rewarded Markov models using regenerative randomization

The First International Workshop on Teaching Performance Analysis of Computer Systems (TeaPACS1) in 2021 came to the conclusion among participants that there is a need to continue and regularly update the discussion on teaching strategies, changing the curriculum, reaching out to the systems community, facing the new generation and their new learning habits, etc.

Performance modeling and analysis of computer systems and networks

Performance and Reliability Analysis of Computer Systems

Mean Value Analysis (MVA) has long been a standard approach for performance analysis of computer systems. While the exact load-dependent MVA algorithm is an efficient technique for computer system performance modeling, it fails to address several features of multi-core platforms. In addition, the load-dependent MVA algorithm suffers from numerical difficulties under heavy load conditions. The goal of our paper is to find an efficient and robust method which is easy to use in practice and also achieves accuracy for performance prediction for multi-core platforms. Our contributions are: We present a flow-equivalent performance model designed specifically to address multi-core computer systems. We identify the influence on the CPU demand of the effects of Dynamic Frequency Scaling (DFS) and Hyper-Threading Technology (HTT). We adopt an approximation technique to estimate resource demands to parameterize the MVA algorithm. We use a modified Conditional MVA (CMVA) algorithm to address the potential numerical instability. To validate the application of our method, we investigate a case study of an e-commerce web server which is equipped with diverse classes of user requests. We show that our method achieves better accuracy compared with other commonly used MVA algorithms.

A new “hybrid” analytic framework, based on the principle of maximum entropy, is used to derive a closed form expression for the queue length distribution of a G/G/1 finite capacity queue. It is shown that “Birth-Death” homogeneous recursions for a single resource queue are special cases of maximum entropy “one-step” transitions which can be applied either in an operational or stochastic context. Furthermore, an “equivalence” relationship is used to analyse two-stage cyclic queueing networks with general service times, and favourable comparisons are made with global balance and approximate results. Numerical examples provide useful information on how critically system behaviour is affected by the distributional form of interarrival and service patterns. Comments on the implication of the work to the performance analysis and aggregation of computer systems are included.

Trace-driven simulation is an important aid in performance analysis of computer systems. Capturing address traces for these simulations is a difficult problem for single processors and particularly for multicomputers. Even when existing trace methods can be used on multicomputers, the amount of collected data typically grows with the number of processors, so I/O and trace storage costs increase. A new technique is presented in this paper which modifies the executable code to dynamically collect the address trace from the user code and analyzes this trace during the execution of the program. This method helps resolve the I/O and storage problems and facilitates parallel analysis of the address trace. If a trace stored on disk is desired, the generated trace information can also be written to files during execution, with a resultant drop in program execution speed. An initial implementation on the Intel iPSC/2 hypercube multicomputer is detailed, and sample simulation results are presented. The effect of this trace collection method on execution time is illustrated.

Queueing theory has become an important subject to computer scientists because it forms the mathematical basis for research in computer system performance evaluation. (See, for example, the Special Issue, September 1978, of Computer Surveys on “Queueing Network Models of Computer System Performance,” and the recent conference on “Applied Probability - Computer Science: The Interface,” which was held in Boca Raton, Florida, in January 1981.) This tutorial will attempt to explain the elements of queueing theory so that the audience will understand (1) the kind of assumptions usually made in the construction of queueing models, (2) the kind of mathematical tools ordinarily used in queueing theory, and (3) the strengths and limitations of queueing theory in the design and analysis of computer systems.

HTB underground gas storage (UGS) is a gas storage facility rebuilt from a depleted gas reservoir. It is located in Changji Xinjiang Province, in northwestern China. With designed working gas volume 45×108m3 and peak working rate 27×106m3 per day, HTB UGS is the biggest UGS constructed in China. Since commissioning in June 2013, HTB UGS has been injecting gas 113.8×108m3, and withdrawing gas 78.2×108m3 until now. In order to analyze its production performance after 7 years operations, we figured out a systematic analysis method. First of all, we built a production database for HTB UGS, collecting daily injection-production data, tubing pressure, and kinds of testing data of the storage. Also, we built a database for all 30 injection-production wells recording their geologic information, drilling information, gas injection-production information, etc. Based on detailed data, we analyzed the operating characteristics of the HTB UGS from 6 aspects: injection-production gas volume, liquid production volume, operating pressure, reservoir connectivity laterally and vertically, sealing of the trap and well-patten spacing. It shows that 1) gas injection and production volume of the storage increases year by year, accounting for 20% working gas volume of UGS in China. 2) Oil and water production are relatively low. The gas is becoming more and more like dry gas. 3) The operating pressure during gas injection almost reaches maximum allowable storage pressure (34 MPa); while during gas withdrawal it reaches 26 MPa, far above the minimum storage pressure (18 MPa). 4) Overall reservoir connectivity is good in the lateral direction and it is disconnected in the vertical direction between E1-2z21 and E1-2z22. 5) The storage is well-sealed and there is no gas leakage. 6) The well-patten covers 80% of the gas storage. To sum up, HTB UGS works well generally from 2013 to 2020, playing an increasingly significant role in summer and winter seasons in China. However, in order to guarantee gas withdrawal in limited time (120 days in winter time in China), well pattern infilling is still needed in layer E1-2z22. The construction of UGS in China started in 1999. At present, 27 gas storage facilities have been built and it is planned to build another 23 to generate more than 400×108m3 working gas volume by the end of 2030. However, a systematic performance analysis method for UGS has not been applied. This paper will offer the reference for production performance analysis of UGS.

Purpose. When solving practical problems that require the creation and further analysis of the model, an important criterion is the labour intensity of modeling. In this regard, the article is aimed at formalizing the modeling process and using the method of step-by-step modeling for the design of technological processes. This approach allows you to design processes and tasks according to the following stages: physical modeling, mathematical modeling, discrete computer modeling and simulation. Methodology. To solve the problem, a methodology of step-by-step modeling is used. The simulation involves 3 stages and uses the decomposition algorithm, i. e. considers the problem from global to detailed. At the first stage of this implementation, the necessary information is collected for the experiment. This information is presented in the form of statistics. In the second stage, further processing takes place, which is performed by checking the compliance of the input data and the process with the question of how this process should be performed. The last stage is the simulation of passages of this fragment, which is represented by a chain of transitions, obtaining statistics of time efficiency of this process, weaknesses of the process and the ability to compare the results obtained during modeling and in the real process, as well as the ability to predict future results and actions. Findings. The technique can be used to study complex technological processes in the enterprise. It allows modeling of complex processes to obtain information about the time efficiency of the technological operation, finding weaknesses in it and patterns in the occurrence of random events that may affect the operation. Using this approach can be very effective in the systems that require constant real-time monitoring, as this tool can be modified by adding sensor kits that will constantly send information to the system or equip an additional system that will provide ready-made information packets. Originality. The method of step-by-step modeling of representation has been improved, which consists in the simultaneous use of physical, mathematical and simulation modeling of complex processes with a set of stages of their implementation. Practical value. The proposed technique is designed for step-by-step modeling of the technological process with the subsequent construction of simulation programming.

We present a distributed active subspace method for training surrogate models of complex physical processes with high-dimensional inputs and function valued outputs. Specifically, we represent the model output with a truncated Karhunen-Lo\`eve (KL) expansion, screen the structure of the input space with respect to each KL mode via the active subspace method, and finally form an overall surrogate model of the output by combining surrogates of individual output KL modes. To ensure scalable computation of the gradients of the output KL modes, needed in active subspace discovery, we rely on adjoint-based gradient computation. The proposed method combines benefits of active subspace methods for input dimension reduction and KL expansions used for spectral representation of the output field. We provide a mathematical framework for the proposed method and conduct an error analysis of the mixed KL active subspace approach. Specifically, we provide an error estimate that quantifies errors due to active subspace projection and truncated KL expansion of the output. We demonstrate the numerical performance of the surrogate modeling approach with an application example from biotransport.

The results obtained from this project will fundamentally change the way we look at computer performance analysis. These results are made possible by the precise definition of a consistent system of measurement with a set of primary units designed specifically for computer performance analysis. This system of units, along with their associated dimensions, allows us to apply the methods of dimensional analysis, based on the Pi Theorem, to define scaling and self-similarity relationships. These relationships reveal new insights into experimental results that otherwise seems only vaguely correlated. Applying the method to cache-miss data revealed scaling relationships that were not seen by those who originally collected the data. Applying dimensional analysis to the performance of parallel numerical algorithms revealed that computational force is a unifying concept for understanding the interaction between hardware and software. The efficiency of these algorithms depends, in a very intimate way, on the balance between hardware forces and software forces. Analysis of five different algorithms showed that performance analysis can be reduced to a study of the differential geometry of the efficiency surface. Each algorithm defines a set of curvilinear coordinates, specific to that algorithm, and different machines follow different paths along the surface depending on the difference in balance between hardware forces and software forces. Two machines with the same balance in forces follow the same path and are self-similar. The most important result from the project is the statement of the Principle of Computational Least Action. This principle follows from the identification of a dynamical system underlying computer performance analysis. Instructions in a computer are modeled as a classical system under the influence of computational forces. Each instruction generates kinetic energy during execution, and the sum of the kinetic energy for all instructions produces a kinetic energy spectrum as a function of time. These spectra look very much like the spectra used by chemists to analyze properties of molecules. Large spikes in the spectra reveal events during execution, like cache misses, that limit performance. The area under the kinetic energy spectrum is the computational action generated by the program. This computational action defines a normed metric space that measures the size of a program in terms of its action norm and the distance between programs in terms of the norm of the difference of their action. This same idea can be applied to a set of programmers writing code and leads to a computational action metric that measures programmer productivity. In both cases, experimental evidence suggests that highly efficient programs and highly productive programmers generate the least computational action.

This paper assesses the state of the art in cost-benefit analyses of computer graphics systems and suggests an approach for developing improved methodology. Cost-benefit analyses are distinguished from analyses of system performance in that the latter is directed at optimizing system performance at a given level of investment, while the former is directed at justifying the investment itself.Computer graphic system design alternatives are first outlined. Then methods of analyzing the performance and costs of computer systems in general and graphic systems in particular are discussed. With this information it is shown how cost-effectiveness analyses may be performed. The next crucial step is to conduct benefit analysis, an ill-defined art. The results of benefit analysis must be combined with cost-effectiveness analysis in order to perform the desired cost-benefit analysis.An experimental methodology is suggested for better performing benefit analyses of computer graphics systems. A more rigorous formulation of the cost-benefit procedure is then outlined. No attempt is made in this report to actually perform such an analysis.

This article is devoted to the analysis of the importance and practical application of simulation and imitation modeling methods for optimal regulation of regional tourism. The article presents scientific approaches to the effective use of resources, visitor flow management, service quality and infrastructure optimization through modeling of complex processes in the tourism sector. Through simulation methods, various scenarios of tourism planning and development are tested and their economic and environmental impact is assessed. The imitation modeling method is shown as an important tool for improving regional tourism strategies. The results of this study create a basis for developing innovative solutions in tourism management and development. В данной статье, посвящена анализу значимости и практическому применению методов имитационного моделирования для оптимального регулирования регионального туризма. В статье представлены научные подходы к эффективному использованию ресурсов, управлению потоками посетителей, качеству обслуживания и оптимизации инфраструктуры посредством моделирования сложных процессов в сфере туризма. С помощью методов имитационного моделирования тестируются различные сценарии планирования и развития туризма, оценивается их экономическое и экологическое воздействие. Метод имитационного моделирования показан как важный инструмент совершенствования региональных стратегий туризма.