Initially, computer systems performance analyses were carried out primarily because of limited resources. Due to ever increasing functional complexity of computational systems and user requirements, performance engineering continues to play a major role in software development.
This book assesses the state of the art in performance engineering. Besides revised chapters drawn from two workshops on performance engineering held in 2000, additional chapters were solicited in order to provide complete coverage of all relevant aspects. The first part is devoted to the relation between software engineering and performance engineering; the second part focuses on the use of models, measures, and tools; finally, case studies with regard to concrete technologies are presented. Researchers, professional software engineers, and advanced students interested in performance analysis will find this book an indispensable source of information and reference.

scihub/10.1007/3-540-45156-0.pdf

zlib/no-category/Dumke, Reiner; Rautenstrauch, Claus; Scholz, André; Schmietendorf, Andreas/[Lecture Notes in Computer Science] Performance Engineering Volume 2047 ||_64927885.pdf

Lecture Notes in Computer Science 2047 Performance Engineering State of the Art and Current Trends

[Lecture Notes in Computer Science] Performance Engineering Volume 2047 ||

Reiner Dumke (editor), Claus Rautenstrauch (editor), Andreas Schmietendorf (editor), Andre Scholz (editor)

Dumke, Reiner; Rautenstrauch, Claus; Scholz, André; Schmietendorf, Andreas

Reiner Dumke, Claus Rautenstrauch, Andreas Schmietendorf, André Scholz

Springer Berlin Heidelberg : Imprint : Springer

Steinkopff. in Springer-Verlag GmbH

Springer London, Limited

Springer Nature

Lecture notes in computer science, 1st ed. 2001, Berlin, Heidelberg, 2001

Lecture notes in computer science -- 2047., Berlin, New York, 2001

1 edition, June 15, 2001

Germany, Germany

1, 2001 may 16

2001, PT, 2001

1, 20030629

sm32586125

Includes bibliographical references and index.

Performance Engineering
Preface
Historical Roots of Performance Engineering
References
Aspects of Performance Engineering - An Overview
1 Motivation
2 Requirements and Aims
3 The Performance Engineering Method
References
Table of Contents
Conception of a Web-Based SPE Development Infrastructure
1 Principles, Models, and Methods of the SPE
2 Evaluation of the Initial Situation According to the Measurement Framework
3 A Static and Open SPE Development Environment
4 Conception of an Intelligent SPE Development Infrastructure
5 Summary
References
Performance and Robustness Engineering and the Role of Automated Software Development
1 Introduction
2 Getting an Early Feedback from the System
2.1 Fast Feedback by Simplification
2.2 Some Counter Examples
2.3 Immediate Feedback by Automation
3.1 The Role of Performance Engineering
3.3 Automated Evaluation of System Properties
System Development with ISG
4.1 ISG: An Overview on the Automated Process Model of ISG
4.2 Application Areas of ISG
4.3 The Feedback Provided by ISG
5 Conclusions and Future Work
References
Performance Engineering of Component-Based Distributed Software Systems
1 Introduction
2 Software Architecture and Performance
2.1 Software Architecture
2.2 Component Interconnection Patterns
2.3 Approach
3 Specifying Component Interconnection Patterns with UML
3.1 Static Modeling of Component Interconnection Patterns
3.2 Dynamic Modeling of Component Interconnection Patterns
4 Performance Annotations of UML-Described Architectures
4.1. Performance Parameters
4.2 Performance Annotations of Components
5 Performance Models of Component Interconnection Patterns
5.1 Synchronous Communication ΠFixed Number of Threads
5.2 Synchronous Communication ΠDynamic Thread Pool
5.3 Asynchronous Communication
6 Conclusions and Future Work
References
Conflicts and Trade-Offs between Software Performance and Maintainability
1 Introduction
2 Industrial Cases
2.1 Billing Gateway (BGw)
Fig. 1. The Billing gateway
2.2 Fraud Control Center (FCC)
2.3 Data Monitoring (DMO)
2.4 Prepaid Service Data Point (SDP)
2.5 Promis Database
3 Studies and Experiences
3.1 Billing Gateway (BGw)
3.2 Fraud Control Center (FCC)
3.3 Data Monitoring (DMO)
3.4 Prepaid Service Data Point (SDP)
3.5 Promise Database
4 Results
5 Discussion
6 Conclusions
References
Performance Engineering on the Basis of Performance Service Levels
Introduction
Database-Based Integrated Information Systems
Performance Engineering
The Service Level Agreement Concept
Performance Service Levels
Process Model
Proof of Service Levels
Outlook
References
Possibilities of Performance Modelling with UML
1 Introduction and Motivation
2 Requirements for a UML-Based Approach
3 Approaches for Performance Engineering Using UML
3.1 Direct Representation of Performance Aspects Using UML
3.2 Expanding UML to Be Able to Handle Performance Aspects
3.3 Combining UML with FDTs on the Example of MSC and SDL
4 Tool Support
4.1 Outline Description of the Function and Use of Tools
4.2 Evaluation of Existing Tools in Table Form
5 An Integrated UML Based Approach
6 Comments and Conclusions
References
Origins of Software Performance Engineering: Highlights and Outstanding Problems
1 Introduction
2 The Evolution of SPE
2.1 Modeling Foundations
2.2 SPE Methods
2.3 SPE Development
2.4 The SPE Process
2.5 Guidelines for Creating Responsive Systems
3 Quantitative Methods for SPE
3.1 Data Requirements
3.2 Performance Models
3.3 Verification and Validation
4 Status and Future of SPE
4.1 Tools
4.2 Performance Models
4.3 Use of SPE
4.4 Concepts for Building High Performance Systems
4.5 SPE Methods
References
Performance Parameters and Context of Use
1 Introduction
Linking Usability-Engineering Principles to Software-Performance Engineering Parameters
3 Analytical Models for Usable Software
References
Using Load Dependent Servers to Reduce the Complexity of Large Client-Server Simulation Models
Introduction
Load Dependent Servers
Obtaining Model Parameters
Obtaining the Frequent States
Estimating the Frequent States Overhead
Fitting Regression Models
Representing Overhead
The Calibration Phase
Calculation of the LDS Capacity
Case Study: Modelling a Server with TPC-C Transactions
The Workload
The Hardware-Software Platform
Obtaining the Frequent States
Estimating the Overhead
The Regression Models
Overhead Representation and Calibration Process
Execution Time
Conclusions and Future Work
References
Performance Evaluation of Mobile Agents: Issues and Approaches
Introduction
Mobile Agents
Different Approaches for MA Performance Analysis
A Framework for Investigating MA Performance
Basic Elements and Application Kernels
A Hierarchical Performance Analysis Approach
Parameterizing Basic Elements and Application Kernels
Implementation and Experimentation with Benchmarks
Experimentation
Micro-Benchmarks
Web Database Micro-Kernels
Conclusions
References
UML-Based Performance Modeling Framework for Component-Based Distributed Systems
1 Introduction
2 Framework Structure
3 UML-Based Modeling Techniques
3.1 Resource Representation
3.2 Workload Representation
3.3 Triggering Properties
3.4 Service Demand Binding
3.5 Network Connections
3.6 Run-Time Configuration
3.7 Creating the AQN Representation
4 Partitioning the Model into Layers
5 Example
6 Conclusions
References
Scenario-Based Performance Evaluation of SDL/MSC-Specified Systems
Introduction
SDL and MSC
Specification and Description Language
Message Sequence Chart
Performance Message Sequence Chart (PMSC)
Model for Performance Evaluation
Definition of Time Constraints
Evaluation Technique
Case Study
Related Work
Conclusion
References
Characterization and Analysis of Software and Computer Systems with Uncertainties and Variabilities
Introduction
Interval Parameters
Definitions and Introduction
Interval Arithmetic
Interval Splitting
Application of Interval-Based Techniques
Modeling Concurrent Applications on a Multicomputer System
Model of an EJB Server Implementation
Histogram Parameters
General Approach: VU-Lists
Sources for VU-Lists
Adaptation of Analysis Techniques
Combination of Intermediate Results
Conclusions and Future Work
References
The Simalytic Modeling Technique
1 Objectives
2 Introduction
3 Background Topics
3.1 Capacity Planning
3.2 Transaction Based Applications
3.3 Client/Server Environments
3.4 Modeling Capacity Projections
3.5 Response Time Modeling
3.6 Modeling Tools
3.7 Platform-Centric
3.8 General Purpose
4 The Simalytic Modeling Methodology
5 Steps to Build a Simalytic Model
5.1 Workload Analysis Phase
5.2 Node Models Phase
5.3 Simulation Model Phase
5.4 Simalytic Model Phase
5.5 Model Analysis
6 Simalytic Model Implementation
6.1 Implementation Example
6.2 Implementation Process
6.3 Workload Analysis Example
6.4 Node Models Example
6.5 Model Results
7 Summary
References
Resource Function Capture for Performance Aspects of Software Components and Sub-Systems
Introduction
Resource Functions
Resource Demand Workbench and Repository
WorkBench Experience
Resource Function Representation
Performance analysis interface
Conclusions
References
Shared Memory Contention and Its Impact on Multi-processor Call Control Throughput
1 Introduction
2 Blocking on Memory Access
3 Model for Memory Contention in SMMPCC
Impact of Shared Memory Contention on SMMPCC
Throughput
4.1 SMMPCC Throughput for Different Numbers of Processors
4.2 SMMPCC Throughput Dependence on the Fractions into Subcall when Contention Occurs
5 Conclusions
References
APPENDIX
Performance and Scalability Models for a Hypergrowth e-Commerce Web Site
1 Introduction
2 Analysis of Daily Traffic
2.1 Bicoastal but Unimodal
2.2 All American!
3 Effective Server Demand
3.1 Modeling Assumptions
3.2 Statistical Approach
4 Choosing Statistical Tools
4.1 Spreadsheet Programming
4.2 The Internet Is Your Friend
5 Planning for Data Collection
5.1 Use It or Lose It
5.2 Saved by the SymbEl
6 Short-Term Capacity
6.1 Multivariate Regression
6.2 Spreadsheet Macros
7 Long-Term Capacity
7.1 Nonlinear Regression
7.2 Doubling Period
8 Upgrade Paths
8.1 Super-Serial Scaling
8.2 Server Scalability Calculation
9 Summary
References
Performance Testing for IP Services and Systems
1 Introduction
2 Performance Testing Process
2.1 Types and Sequence of Performance Tests
2.2 Performance Testing Process Flow
Performance Test Planning Activities. The performance test planning activities shown in Figure 1 are described below.
Performance Test Execution and Analysis Activities
2.3 Ensuring a Successful Performance Testing Effort
3 Using Stress Test Results to Drive Performance Modeling:
A Case Study in fiGray-Boxfl Vendor Analysis
3.1 Stress Testing Environment
Stress Testing Results
3.3 Stress Testing Theory
3.4 Stress Test Results, Revisited
3.5 Stress Testing Model
3.6 Model Parameterization
3.7 Vendor Follow-Up
4 Conclusions
References
Performance Modelling of Interaction Protocols in Soft Real-Time Design Architectures
1 Introduction
2 Basic Interaction Protocols
2.1 Channel
2.2 Pool
2.3 Signal
2.4 Constant
3 Protocol Extensions
4 Protocol Taxonomy
4.1 Simple Interaction Protocols
4.2 Composite and Bi-Directional Interaction Protocols
5 Performance Modelling of Interaction Protocols
5.1 Single server Semaphore queue (SsSq)
5.2 Multi-server Semaphore queue (MsSq)
5.3 Resolution Considerations Using Semaphore Queues
6 Conclusions and Future Work
References
A Performance Engineering Case Study: Software Retrieval System
Introduction
System Specification
Modelling the Software Retrieval System
Modelling Using UML
Modelling Using Stochastic Petri Nets
Performance Analysis of the SRS
Performance Assumptions
Performance Results
Concluding Remarks
References
Performance Management of SAP® Solutions
1 Introduction
1.1 A Case Study
1.2 Performance Tuning
1.3 Excursus: From SAP R/3® to mySap.com®
2 Parallel Processing ΠRequirements on System Configuration
2.1 SAP Client/Server Architecture
2.2 Scalability
2.3 Benchmarks
2.4 Sizing Methods
3 Parallel Processing ΠLocking Constrains
3.1 ATP Check
3.2 Database Locks on Number Range Objects
4 Skills, Methods, and Service Strategies in SAP Performance
References
Author Index

<p><P>Initially, computer systems performance analyses were carried out primarily because of limited resources. Due to ever increasing functional complexity of computational systems and user requirements, performance engineering continues to play a major role in software development.<br>This book assesses the state of the art in performance engineering. Besides revised chapters drawn from two workshops on performance engineering held in 2000, additional chapters were solicited in order to provide complete coverage of all relevant aspects. The first part is devoted to the relation between software engineering and performance engineering; the second part focuses on the use of models, measures, and tools; finally, case studies with regard to concrete technologies are presented. Researchers, professional software engineers, and advanced students interested in performance analysis will find this book an indispensable source of information and reference.</p>

Lecture Notes in Computer Science
Erscheinungsdatum: 23.05.2001

2014-12-23