In the design, implementation, and operational planning of computer and c- municationsystems, manyquestionsregardingthe desiredcapacityandspeed of (sub)systems have to be answered.At this point, performance anddependability evaluation techniques can be of great help. With these techniques, design de- sions can be prepared using advanced methods to construct appropriatemodels, toparameterisethesemodels, andtosolvethem.Theapplicationofabroadsp- trum of such methods and techniques is currently supported by tools (mostly software, but sometimes partly hardware as well). Such tools enable system - signers and engineers to construct their models in a ?exible and modular way using high-level application-oriented modelling languages, to solve their models with a variety of techniques and to exercise parametric studies at ease. The goal of the 11th International Conference on Modelling Tools and Techniques for Computer and Communication System Perf- mance Evaluation (\TOOLS 2000") was to further develop the theory and technologyfortool-basedperformanceanddependability evaluationofcomputer and communication systems. Important themes included software tools, eval- tion techniques, measurement-based tools and techniques, performance and - pendabilityevaluationtechniquesbasedonformalmethods, casestudiesshowing the role of evaluation in the design of systems, and application studies in the area of centralised and distributed computer systems. Previous conferences in this series were held over the past 15 years in Paris (1984), Sophia Antipolis (1985), Paris(1987), Palma de Mallorca (1988), Torino (1991), Edinburgh (1992), Wien (1994), Heidelberg (1995), Saint Malo (1997), and Palma de Mallorca (1998). The proceedings of the latter four conferences also appeared in the series Lecture Notes in Computer Science (Volumes 794, 977, 1245, and 1469, respectively).

lgrsnf/2105.pdf

scihub/10.1007/3-540-46429-8.pdf

Computer performance evaluation : modelling techniques and tools : 11th international conference, TOOLS 2000, Schaumburg, IL, USA, March 27-31, 2000 : proceedings

International Conference on Modeling Techniques and Tools for Computer Performance Evaluation

Boudewijn R. Haverkort (editor), Henrik C. Bohnenkamp (editor), Connie U. Smith (editor)

TOOLS 2000 (11th 2000 Schaumburg, Ill.)

Springer Spektrum. in Springer-Verlag GmbH

Steinkopff. in Springer-Verlag GmbH

Lecture notes in computer science -- 1786, Berlin, New York, Germany, 2000

Springer Nature, Berlin, Heidelberg, 2003

1 edition, April 13, 2000

Germany, Germany

sm38858160

{"isbns":["3540672605","9783540672609"],"last_page":404,"publisher":"Springer","source":"libgen_rs"}

Lecture Notes in Computer Science
Computer Performance Evaluation
Preface
Scientific Organisation
Organisation
Table of Contents
Performance Evaluation with Heavy Tailed Distributions (Extended Abstract)
Introduction
Evidence
Implications of Heavy Tails
Taking Advantage of Heavy Tails
Two Important Properties
Exploiting the Heavy Tail Properties
Layered Modeling of Hardware and Software, with Application to a LAN Extension Router
Introduction
Application of RMA to a LAN Extension Router
A More Detailed Description of the LAN Extension Router
Derivation of the Performance Model
Step 1: Obtaining Sub-models from Scenarios
Step 2: Merge Sub-Models to Create a Single Model\ of the System
Step 3: Adding Behavioral Detail to Model
Step 4: Simplifying the Model
Model Parameters and Validation
Parameter Gathering
Model Validation Experiments
Model Validation Results
Conclusions
Object Allocation for Distributed Applications with Complex Workloads
Introduction
Performance Requirements and Capacity Constraints
Capacity Constraints, Workloads, and Performance Requirements
Satisfaction Function
Finding the Maximum Mean Response Times for a Given Configuration
Differences Between Medium and High Workload Conditions
The Programming Model
Implementing the Programming Model
Finding Configurations that Maximize Satisfaction
Case Study – A Multi Tier Distributed Application
Summary and Conclusions
References
A Set of Tools for Traffic Modeling, Analysis and Experimentation
1 Introduction
2 Traffic Characterization
3 Descriptors Computation for Markov Chains
4 Modules for Traffic Modeling, Analysis and Experimentation
5 Examples
5.1 Traffic Models
5.2 Results
6 Summary
References
Queueing Analysis of Pools in Soft Real-Time Systems
Introduction
Semaphore Queues
Pools
Numerical Examples
Conclusions and Open Problems
Xaba: Exact, Approximate, and Asymptotic Solvers for Multi-class Closed Queueing Networks*
1 Introduction
2 Solution Methods
2.1 Exact Solution Methods
2.2 Approximate Solution Methods
2.3 Asymptotic Bounds Methods
2.4 Evaluation Methodology
3 Xaba
3.1 The Architecture
3.2 Functionalities
4 An Evaluation Study
4.1 The System
4.2 The Model
4.3 Results
5 Conclusions
References
Decomposition of General Tandem Queueing Networks with MMPP Input
Introduction
Traffic Descriptors
Two-State MMPPs
Two-State SMPs
The MMPP(2)/G/1(/K)Queue and Its Output
The Output of the MMPP(2)/G/1/1Queue
Approximation of the Output of the MMPP(2)/G/1(/K>1)Queue
Conversion of SMP(2)to MMPP(2)
Approximate Analysis of Tandem Networks \ with and without Customer Loss
Conclusions
Exact and Approximate Solutions for a Class of Infinite Markovian Models
Introduction
The Basic Solution
Ergodicity
Normalization
The Approximate Solutions
The Krylov Subspaces Procedure
Finding a Proper Value for m
Getting the Approximate Solution
An Example
Conclusions
Performance Evaluation of a Distributed Enterprise Data Mining System
Introduction
Kensington EJB Server
Enterprise JavaBeans
EJB Server Implementation
Method Call Execution
An Extended Queueing Model
Blocking Time
The FES and Whole Model
Instance-Active and Client-Blocking Probabilities
Results
Conclusion
Minimization of the Impact of Subscriber Mobility on the Resources of a GSM Network
Introduction
Exact Description of the Problem
Mathematical Model of the Problem
Optimization of the Network Topology
Conclusions
Optimal Allocation Policies for Mobile Agents
Introduction
Finite Horizon Optimization
Infinite Horizon Optimization
Numerical Results
Analysis of a Special Case
Conclusion
Iterative Analysis of Markov Regenerative Models
Introduction
The Conventional Analysis Procedure
Uniformization for General Distributions
An Iterative Algorithm for Fill-in Avoidance
Complexity Analysis
Examples
Conclusions
Analysis and Evaluation of Non-Markovian Stochastic Petri Nets
Introduction
Introducing Petri Nets and Preemption Policies
Discrete-Time Approach
General Solution
The Algorithm
Numerical Results
Conclusions
Petri Net Modelling and Performability Evaluation with TimeNET 3.0
Introduction
Supported Net Classes and Analysis Methods
Markov Regenerative Stochastic Petri Nets
Coloured Petri Nets for Manufacturing Systems
Agnes -- The New TimeNET Interface
Tool Usage
Net Classes and Modules
A Manufacturing System Application Example
Structural Model
Production Sequence Model
Performability Evaluation of the Example
Conclusions
Integrating Synchronization with Priority into a Kronecker Representation*
Introduction/Motivation
Basic Definitions
Kronecker Expression with Prioritized Synchronization
Analysis Algorithms
Analysis of an Example and Tools Integration
Conclusions
Implementation of Importance Splitting Techniques in Stochastic Petri Net Package
Introduction
SPNP: Stochastic Petri Net Package
Importance Splitting Techniques
Basic Idea of Importance Splitting
Splitting
RESTART
Implementation in SPNP
Numerical Illustrations
Dual-Tank Example
Reader and Writer Example
Conclusion
FluidSim: A Tool to Simulate Fluid Models of High-Speed Networks
Introduction
Dynamics of a Fluid Model
Fluid Model of a Communication Network
Simulation of a Fluid Network
Event Classes
Event Handling
Events Related to the Sources
Events Related to the Buffers
Events Related to the Switching Matrices
Events Related to the Communication Links
Implementation of the Fluid Simulator
The Simulator Kernel
Network Objects
Configuration Interface
Examples
Comparison with an Analytical Model
Comparison with a Cell-Level Simulation Tool
Conclusions
Exploiting Modal Logic to Express Performance Measures
Introduction
PEPA
The PEPA Reward Language
Implementation
Case Study: Self-Checking Distributed Computation
Related Work
Derivation of Petri Net Performance Models from UML Specifications of Communications Software
Introduction
Modelling ABP in the Unified Modelling Language
The Alternating Bit Protocol
The Unified Modelling Language
Using UML for Performance Modelling
Exploiting Implementation Diagrams
Direct Simulation of Sequence Diagrams
Deriving a Petri Net Model of ABP
Petri Nets
Use Case and Workloads
A UML Combined State and Collaboration Model
The Corresponding GSPN Model
Numerical Results
Current Work
Conclusions
Scalability of a Workstation Cluster Architecture for Video-on-Demand Applications
Introduction
Scalability and System Requirements
Experimental Approach
Architecture
The Cluster Architecture
Scalability
Scaling Functions
Hierarchical Scaling
Workload
Technology
The Forecasting Problem
Technology Reference Points2
Measurement Study
Streaming Capacity in Play Mode
Effect of User Interaction on Streaming Capacity
Effect of Variable Bit Rate
Node-Level Analysis (Cluster)
Structural Analysis
Device-Level Projections
Projections for Node-Level Scalability
System-Level Scalability
System Streaming Capacity
Node Capacity
Capacity Increments
System Capacity Limits
On-Line Storage Requirements
Conclusion
Multilayer Online-Monitoring for Hybrid DSM Systems on Top of PC Clusters with a SMiLE
Motivation
Shared Memory on Clusters and Its Problems
The SMiLE Monitoring System
Layers of Data Acquisition
Hardware Monitoring
Status Information from the Hybrid HW/SW-DSM System
Instrumenting the Synchronization Primitives
Programming Model and Environment Specific Information
Unifying to a Single Abstract Model
On-Line Monitoring with OMIS
Bringing It All Together
Current Status and Future Work
Related Work
Conclusions
Design and Implementation of a Tool for Measuring the Performance of Complex E-commerce Sites*
1 Introduction
2 Architecture of E-commerce Sites
2.1 Implementing E-commerce Services
2.2 Performance and Congestion in E-commerce Sites
2.3 Characteristics of E-commerce Workloads
3 Measuring the Performance of E-commerce Server
3.1 Measurement Approach
3.2 Architecture of the Tool
3.3 The Interface of PROFIT
4 Case Studies
4.1 Experimental Setup
4.2 Workload Generation
4.3 Example 1: Overhead of Secure Connections
4.4 Example 2: US vs. International Customers
4.5 Example 3: Comparing Apache Modules to CGI
4.6 Example 4: Evaluating Store Scalability
5 Conclusions and Future Work
References
FiFiQueues: Fixed-Point Analysis of Queueing Networks with Finite-Buffer Stations
1 Introduction
2 Tool Description
2.1 Model Specification
2.2 Evaluation Results
3 Employed Algorithms
3.1 Traffic Analysis
3.2 Node Analysis
4 Outlook
References
Galileo: A Tool for Dynamic Fault Tree Analysis
Introduction
A Mission Avionic System Example
Obtaining Galileo
Current Work
M ̈obius: An Extensible Tool for Performance and Dependability Modeling *
Introduction
Möbius Framework
Möbius Tool
Future Directions
MRMSolve: A Tool for Transient Analysis of Large Markov Reward Models*
Introduction.
The Structure of MRMSolve.
Graphical User Interface.
Data Representation.
Application Example.
Future Plans.
Calculation and Use of Peaking Factors for Remote Terminal Emulation
Remote Terminal Emulation
Benchmark Test Workload Definition
Calculating Peaking Factors
Using the Peaking Factor
Reliability and Performability Modeling Using SHARPE 2000
1 Introduction
2 SHARPE
2.1 SHARPE Menu of Model Types
3 Recent Additions to SHARPE Engine
4 Graphical User Interface
References
The Simalytic Modeling Technique: Overview for Application Performance Analysis
Introduction
Simalytic Modeling Methodology
Conclusion
SPNP: Stochastic Petri Nets. Version 6.0
Introduction
Solution Methods
Analytic-Numeric Methods
Simulation
iSPN 6.0
SREPT: Software Reliability Estimation and Prediction Tool
Introduction and Motivation
Architecture of SREPT
STEADY - A Tool for Predicting Performance of Parallel DBMSs
STEADY
Comparison with Process Algebra
Conclusions
The TANGRAM-II Environment
1 Introduction
2 Model Specification
3 Model Solution and Traffic Modeling
References
Implementing Symbolic Models for Value Passing in TwoTowers
Introduction
Implementing Value Passing in TwoTowers
Advantages of Value Passing
Monitoring Application-Level VoIP Grade of Service at a PSTN-IP Network Gateway
1 Summary
÷Prof-SDL: A Tool for Identifying Performance Bottlenecks of Software Architectural Design
Extended Abstract
Author Index

<p><P>This book constitutes the refereed proceedings of the 11th International Conference on Modelling Tools and Techniques for Computer Communication System Performance Evaluation, TOOLS 2000, held in Schaumburg, IL, USA in March 2000. The 21 revised full papers presented were carefully reviewed and selected from a total of 49 submissions. Also included are 15 tool descriptions and one invited paper. The papers are organized in topical sections on queueing network models, optimization in mobile networks, stochastic Petri nets, simulation, formal methods and performance evaluation, and measurement tools and applications.</p>

2015-05-26