You are previewing HP-UX 11i Tuning and Performance.
O'Reilly logo
HP-UX 11i Tuning and Performance

Book Description

Maximize the performance and value of any HP-UX 11i system

HP-UX 11i Tuning and Performance takes the guesswork out of maximizing HP-UX 11i performance—so you can quickly maximize the value of any HP-UX system in any environment. Leading HP experts introduce a comprehensive, systematic performance management methodology that addresses every stage in the lifecycle of an HP-UX system.

You’ll find specific metrics, symptoms and solutions for every key element of an HP-UX 11i system, including hardware, the OS platform, and application development. From start to finish, the emphasis is on practical results—and on understanding the real-world cost/benefit tradeoffs that are central to every tuning project. Coverage includes:

  • A comprehensive optimization methodology, direct from HP’s performance labs

  • New optimization opportunities and techniques made possible by HP-UX 11i

  • All-new chapters on network performance and Java performance

  • Rules of thumb for more cost-effective performance management

  • Instrumenting HP-UX 11i systems to accurately reflect reality

  • Choosing the right performance metrics, and using them correctly

  • Key performance tradeoffs in both uniprocessor and SMP environments

  • Practical techniques for resolving CPU, memory, and disk bottlenecks

  • How to make the most of HP-UX specific and generic UNIX performance tools

  • Performance enhancements available through compiler optimization and application profiling

  • Extensive examples from actual HP performance consulting assignments

  • Whether you’re an HP-UX 11i sysadmin, netadmin, developer, capacity planner, or performance specialist, this is your complete guide to getting all the performance you paid for!

    Table of Contents

    1. Copyright
    2. Hewlett-Packard® Professional Books
    3. Preface
    4. About the Authors
    5. Acknowledgments
    6. Introduction to Performance Management
      1. Application Developer's Perspective
      2. System Administrator's Perspective
      3. Total System Resource Perspective
      4. Rules of Performance Tuning
    7. Performance Management Tasks
      1. Workload Estimation
      2. Benchmarking
      3. Performance Characterization
      4. Performance Forecasting or Prediction
      5. Application Optimization
      6. Capacity Planning
      7. Performance Problem Resolution
    8. A Performance Management Methodology
      1. Assessment
      2. Measurement
      3. Interpretation and Analysis
      4. Identifying Bottlenecks
      5. Tuning or Upgrading
    9. Kernel Instrumentation and Performance Metrics
      1. Approaches to Measuring
      2. Kernel Instrumentation and Measurement Interface
      3. Performance Metrics Categories
      4. Summary
    10. Survey of Unix Performance Tools
      1. Choosing the Right Tools
      2. Multipurpose Diagnostic Tools
      3. CPU Diagnostic Tools
      4. Disk Diagnostic Tools
      5. Memory Diagnostic Tools
      6. Performance Characterization and Prediction Tools
      7. Process Accounting Tools
      8. Application Optimization Tools
      9. Network Diagnostic Tools
      10. Resource Management Tools
    11. Hardware Performance Issues
      1. Terminology
      2. Processor Characteristics
      3. Multi-Processing
      4. Cache Memory Performance Issues
      5. Main Memory Performance Issues
      6. I/O Performance Issues
      7. PA-RISC Systems
      8. Itanium® Systems
    12. CPU Bottlenecks
      1. Processes and Threads
      2. Scheduling
      3. System Partitioning
      4. Traps and Protection Violations
      5. Global CPU Metrics
      6. Typical Metric Values
      7. Symptoms of a CPU Bottleneck
      8. CPU Use and Performance Tools
      9. Tuning CPU Bottlenecks
    13. Memory Bottlenecks
      1. Virtual Address Space
      2. Memory Organization
      3. Paging Control
      4. File Access
      5. Process and Thread Execution
      6. Malloc'ed Memory
      7. System V Shared Memory
      8. Shared Libraries
      9. Memory Management Policies
      10. Sizing Memory and the Swap Area
      11. Memory Metrics
      12. Types of Memory Management Bottlenecks
      13. Expensive System Calls
      14. Tuning Memory Bottlenecks
      15. Memory-Related Tuneable Parameters
    14. Disk Bottlenecks
      1. Disk Hardware Descriptions
      2. Review of Disk I/O Concepts
      3. Logical Volume Manager Concepts
      4. VxVM Volume Manager
      5. Shared-Bus Access to Disks
      6. File Systems and the Kernel
      7. File System Types
      8. Disk Metrics
      9. Types of Disk Bottlenecks
      10. Expensive System Calls
      11. Tuning Disk Bottlenecks
      12. Database Issues
      13. Disk-Related Tuneable Parameters
    15. Network Bottlenecks
      1. Networking Hardware Descriptions
      2. Review of Networking Concepts
      3. Networking Access
      4. Networked File System
      5. Clusters
      6. Network Metrics
      7. Types of Network Bottlenecks
      8. Expensive System Calls
      9. Tuning Network Bottlenecks
      10. Network-Related Tuneable Parameters
      11. Web Server Tuning Issues
      12. Database Server Tuning Issues
    16. Compiler Performance Tuning
      1. Compilers and Optimization
      2. Optimization Levels
      3. Compiling for a Target Runtime Environment
      4. Finer Control Over Optimization
      5. Linker Optimizations
      6. Profile-based Optimization
      7. Specific Options for Fortran and COBOL
      8. Why Does Optimization “Break” Applications?
      9. Debugging Optimization Problems
      10. Porting Applications
      11. Code to Demonstrate Optimization Effects
    17. Java Run-time Performance Tuning
      1. Overview
      2. Performance Issues
      3. Versions on HP-UX
      4. HP-UX Tuneable Parameters and Java
      5. Using Large Java Heap Sizes
      6. Performance Options
    18. Designing Applications for Performance
      1. Tips for Application Design
      2. Shared Versus Archive Libraries
      3. Archive Libraries
      4. Choosing an Inter-Process Communication (IPC) Mechanism
      5. Trade-offs with Network System Calls
      6. Instrumenting an Application for Performance Monitoring
    19. Application Profiling
      1. Caliper
      2. CXperf
      3. gprof
      4. Puma
      5. HPjmeter
    20. Performance Tools Alphabetical Reference
      1. System V Tools
      2. Berkeley Tools
      3. HP-UX Tools
      4. Other Tools
      5. Headings in the Summary Chart
      6. Tool Reference Tables
    21. HP-UX Version Naming Reference
    22. Dynamically Tuneable Parameters
    23. Index