Table of Contents

Performance Tuning for Linux® Servers
By Sandra K. Johnson, Gerrit Huizenga, Badari Pulavarty
...............................................
Publisher: Prentice Hall PTR
Pub Date: May 27, 2005
ISBN: 0-13-144753-X
Pages: 576

Table of Contents  | Index

		Copyright
		IBM Press
			On Demand Computing Books
			WebSphere Books
			More Books from IBM Press
			DB2® Books
		Preface
		Acknowledgments
		About the Editors
		About the Contributors
		Part I.  Linux Overview
			Chapter ONE.  Linux Installation Issues
				Introduction
				Preinstallation Planning
				Configurable 2.6 Kernel Features
				Linux Logging Facility
				System Initialization: BSD Versus System V Initialization
				Summary
				References
			Chapter TWO.  Kernel Overview
				Introduction
				The Evolution of Linux
				Linux Kernel Architecture
				Process Management
				Interprocess Communications
				The Linux Symmetrical Multiprocessing (SMP) Model
				File Systems
				New Features in Linux 2.6
				Summary
				References
			Chapter THREE.  Overview of Server Architectures
				Introduction
				Linux Servers
				Processors and Multiprocessing
				Memory
				I/O
				Linux Enterprise Servers
				Linux Clusters
				Examples of Server Systems
				Summary
		Part II.  Performance Analysis Tools
			Chapter FOUR.  System Performance Monitoring
				Introduction
				Background on Linux and Performance Analysis
				CPU Utilization
				Memory Utilization
				I/O Utilization
				Network Utilization
				Summary
				References
			Chapter FIVE.  System Trace Tools
				Introduction
				Requirements for System Tracing
				The top Utility
				strace
				OProfile
				Performance Inspector
				Summary
				References
			Chapter SIX.  Benchmarks as an Aid to Understanding Workload Performance
				Introduction
				Benchmarking to Improve Your Workload
				What Types of Benchmarks Are There?
				Microbenchmarks
				Web Server Benchmarks
				Summary
		Part III.  System Tuning
			Chapter SEVEN.  System Performance Principles and Strategy: A Benchmarking Methodology Case Study
				Introduction
				Performance Evaluation Methodologies
				Benchmarking Methodology Case Study
				Analysis Methodology
				Benchmarks
				Summary
				Acknowledgments
				References
			Chapter EIGHT.  Scheduler Tuning
				Introduction
				Single-Processor Systems
				Symmetric Multiprocessing (SMP)
				Non-Uniform Memory Access (NUMA)
				Symmetric Multithreading (SMT)
				The 2.6 Linux Scheduler
				Load Balancing
				Tunable Parts of the Scheduler
				Summary
				References
			Chapter NINE.  The Linux Virtual MemoryPerformance Implications
				Introduction
				Memory and Address Space
				High-Memory Support
				Paging and Swapping
				The Linux Page Tables
				New Features in Linux 2.6
				Summary
				References
			Chapter TEN.  I/O SubsystemsPerformance Implications
				Introduction
				I/O Scheduling and the Block I/O (BIO) Layer
				Read and Write Request Batches
				Read Anticipation Heuristic
				I/O Components That Affect Performance
				Addressing an I/O Device
				Summary
				References
			Chapter ELEVEN.  File System Tuning
				Introduction
				File System Fundamentals
				Journaled File Systems
				Disks Factor into File System Performance
				Fragmenting a File System
				File Synchronization
				bdflush Parameters
				Asynchronous Input and Output
				Raw Disk I/O
				Ext2 and Ext3
				ReiserFS
				Journaled File System (JFS)
				Next-Generation File System (XFS)
				Summary
				References
			Chapter TWELVE.  Network Tuning
				Introduction
				The Network Protocol Stack
				Kernel Parameter Tuning Mechanisms
				Kernel Auto Tuning
				Core Kernel Parameter Descriptions
				TCP/IPv4 Protocol Kernel Parameters
				Summary
				References
			Chapter THIRTEEN.  Interprocess Communication
				Introduction
				What Is Interprocess Communication?
				Linux SysV IPC Resources and the ipcs Command
				Semaphore Parameters
				Message Queue Parameters
				Shared Memory Segment Parameters
				Dynamically Modifying the Configurable IPC Parameters
				Configuring IPC Parameters Statically
				Pipes
				Summary
			Chapter FOURTEEN.  Code Tuning
				Introduction
				General Principles
				Profiling to Understand the Application
				Compiler Options as Tuning Tools
				Code Tuning
				Algorithm: Achieving Performance Through Design Choices
				Summary
		Part IV.  Performance Characterization of Linux Server Applications
			Chapter FIFTEEN.  Web Servers
				Introduction
				HTTP Requests and Responses
				Network Behavior of a Web Server
				Anatomy of a Web Server Transaction
				Different Models of Web Servers
				Tuning Web Servers
				Summary
				References
			Chapter SIXTEEN.  File and Print Servers
				Introduction
				Types of Dedicated Network Storage Servers
				Optimizing the Performance of Network Storage
				References
			Chapter SEVENTEEN.  Database Servers
				Introduction
				Overview of Database Architectures
				Database Tuning Areas to Consider
				Process Management
				Memory Management
				I/O Management
				Summary
			Chapter EIGHTEEN.  Application Servers
				Introduction
				The Application Server Defined
				Java, J2EE, and Application Servers
				Performance Characterization of Application Servers
				Improving Performance and High Availability
				Summary
				References
		Part V.  Tuning Case Studies
			Chapter NINETEEN.  Case Study: Tuning the I/O Schedulers in Linux 2.6
				Introduction
				Benchmark Environment and Workload Profiles
				I/O Schedulers and Performance
				Single-CPU Single-Disk Setup
				8-Way RAID-5 Setup
				16-Way RAID-0 Setup
				AS Sequential Read Performance
				AS Versus Deadline Performance
				CFQ Performance
				Summary
				References
			Chapter TWENTY.  Case Study: File System Tuning
				Introduction
				Analyzing File Layout
				Tuning File Systems
				Measuring I/O
				Summary
				References
			Chapter TWENTY-ONE.  Case Study: Network Performance on Linux
				Introduction
				Benchmarks Used in This Case Study
				Enhancements in the Linux 2.4 and 2.6 Kernels
				Case Study
				Summary
				References
			Chapter TWENTY-TWO.  Case Study: Commercial Workload Tuning
				Introduction
				Overview of Commercial Workload Tuning
				Standard Commercial Workload Model for J2EE
				Our Commercial Workload Model: Stock Trading
				The Performance Analysis Exercise
				Summary
				References
			Appendix A.  Tuning Kernel Parameters
				Introduction
				The sysctl Interface
				The procfs Interface
				sysfs (Linux Kernel 2.6 Only)
				General Kernel Parameters
				Virtual Memory
				File System
				Network
		Index