Understanding IBM eServer pSeries Performance and Sizing

Book description

This IBM Redbooks publication is an update to the successful first edition of Understanding IBM RS/6000 Performance and Sizing, SG24-4810 that was published in 1997. It gives a broad overview of IBM RS/6000 and IBM eServer pSeries performance and sizing.

Contained in this book is a close-up, performance related view of the different hardware architectures IBM offers in its RS/6000 and eServer pSeries systems, including system, processor, memory, storage, and network architectures. One chapter is dedicated to general sizing rules for a number of environments such as database sizing, IBM HTTP server sizing, Net.Commerce sizing, and Lotus Domino sizing. The reader will also find a description of the industry benchmarks that are performed on IBM systems as well as an overview of AIX performance tools.

This book is intended to be a handbook for anyone who wants to gain a deeper understanding of IBM RS/6000 and IBM eServer pSeries architectures from a performance perspective.

Table of contents

  1. Preface
    1. The team that wrote this redbook
    2. Comments welcome
  2. Chapter 1: Introduction
  3. Chapter 2: Background
    1. Performance of processors
    2. Hardware architectures
      1. RISC/CISC concepts
      2. Superscalar architecture: pipeline and parallelism
      3. Memory management (1/2)
      4. Memory management (2/2)
      5. PCI
      6. MP implementation specifics
      7. NUMA
      8. Logical partitioning (LPAR)
    3. AIX kernel
      1. Description
      2. Executable file formats
      3. Kernel and user mode
      4. I/O
      5. Context/Thread switches
      6. Virtual address space
      7. Demand paging
      8. Kernel scalability enhancements
      9. References
    4. 64-bit architecture
      1. Concepts
      2. Addressability
      3. Advantages of 64-bit architecture
      4. Performance of 64-bit architecture
      5. Software considerations for 64-bit architecture
      6. 64-bit operating system capabilities
  4. Chapter 3: IBM RS/6000 and IBM pSeries architectures
    1. POWER2 Super Chip
    2. POWER3
      1. POWER3 execution core
      2. Memory access section
      3. POWER 3 II chip
    3. PowerPC
      1. PowerPC 604 and 604e
      2. Differences between 604 and 604e processors
      3. RS64 II processor (1/2)
      4. RS64 II processor (2/2)
      5. RS64 III processor (1/3)
      6. RS64 III processor (2/3)
      7. RS64 III processor (3/3)
      8. POWER4
  5. Chapter 4: IBM RS/6000 and IBM pSeries products
    1. Symmetrical Multiprocessor (SMP)
      1. Migrating to SMP
      2. Symmetrical Multiprocessor (SMP) concepts and architecture (1/2)
      3. Symmetrical Multiprocessor (SMP) concepts and architecture (2/2)
      4. Software (1/2)
      5. Software (2/2)
      6. Scaling
      7. References
    2. Scalable POWERparallel (SP)
      1. Parallel architecture
      2. IBM SP (Scalable POWERparallel) system
      3. SP switch performance (1/2)
      4. SP switch performance (2/2)
      5. Shared disk components of Parallel System Support Programs
      6. Sizing and configuring a control workstation
      7. Sizing and configuring an SP system (1/4)
      8. Sizing and configuring an SP system (2/4)
      9. Sizing and configuring an SP system (3/4)
      10. Sizing and configuring an SP system (4/4)
      11. Resources
  6. Chapter 5: Hardware
    1. Processors
    2. Memory
      1. Cache memory
      2. Addressing considerations
      3. Memory cycles
      4. Uniprocessor vs. symmetric multiprocessor memory cycles
      5. Miss rate penalty
      6. Effect of L2 cache
      7. Effect of processor speed
    3. Storage
      1. Performance view
      2. Levels of storage
      3. How an I/O request is processed
      4. How a disk works
      5. SCSI technology
      6. Serial Storage Architecture (SSA)
      7. RAID levels overview and performance considerations
      8. IBM Enterprise Storage Server (ESS)
      9. Logical Volume Manager (LVM) concepts (1/2)
      10. Logical Volume Manager (LVM) concepts (2/2)
      11. Raw logical volumes versus Journaled File Systems (JFS)
    4. Asynchronous Communication adapters
      1. Terms used in serial communication
      2. Flow control
      3. Asynchronous adapter overview
      4. Evaluating asynchronous communications options
      5. Product selection considerations
      6. Topology considerations
    5. LAN/WAN Adapters
      1. Ethernet
      2. Token Ring
      3. Fibre Channel
      4. ATM
      5. General network tuning recommendations
    6. Graphics accelerators
      1. Currently available RS/6000 graphics accelerators
      2. IBM’s graphics workstations
      3. Graphics APIs - The “softer side of things”
      4. Graphics accelerator positioning
      5. References
    7. Network Station
      1. Network Station memory
      2. Boot server performance
      3. Boot performance considerations
      4. Application performance considerations
      5. Using CDE with the Network Station
      6. Performance summary
  7. Chapter 6: Benchmarks
    1. System Performance Evaluation Corporation (SPEC)
      1. SPEC CPU2000
      2. SPEC JVM98
      3. SPEC SFS97
      4. SPEC web99
      5. Reference
    2. Graphics Performance Characterization (GPC) Committee
      1. SPECviewperf
    3. Transaction oriented benchmarks
      1. TPC-C
      2. TPC-D
      3. TPC-H
      4. TPR-R
      5. TPC-W
    4. ROLTP
    5. LINPACK
      1. Metrics
      2. Usage
      3. Reference
    6. NotesBench benchmark
      1. NotesBench test
      2. NotesBench test scenario
      3. Metrics and how to read them
      4. Usage
      5. Conclusion
      6. References
  8. Chapter 7: Sizing
    1. General sizing concepts
      1. Guidelines
      2. Concepts (1/2)
      3. Concepts (2/2)
      4. Using AIX Workload Manager (WLM)
      5. Resources
    2. Multiuser system sizing
      1. Multiuser environment
      2. Workload balancing
      3. General sizing considerations
      4. Resources
    3. File server sizing
      1. NFS sizing (1/2)
      2. NFS sizing (2/2)
      3. AIX Fast Connect sizing
      4. Client/Server sizing
      5. General sizing considerations
      6. Resources
    4. Database sizing
      1. Database environment
      2. Transaction processing monitor environment
      3. Sizing RDBMS (1/2)
      4. Sizing RDBMS (2/2)
      5. Resources
    5. Web server sizing
      1. Introduction
      2. Sizing preparation
      3. Sizing factors
      4. Web server performance
      5. Sizing IBM HTTP Server
      6. Sizing WebSphere Application Server (1/2)
      7. Sizing WebSphere Application Server (2/2)
      8. Sizing Net.Commerce
      9. Resources
    6. Lotus Domino Server sizing
      1. Estimate the workload
      2. Processor sizing
      3. Memory sizing
      4. Disk sizing
      5. Example
      6. Conclusion
      7. Resources
  9. Chapter 8: Performance tools
    1. AIX performance tools and commands
      1. Commands viewed by filesets
      2. Commands viewed by system resource
      3. Command descriptions (1/6)
      4. Command descriptions (2/6)
      5. Command descriptions (3/6)
      6. Command descriptions (4/6)
      7. Command descriptions (5/6)
      8. Command descriptions (6/6)
      9. References
    2. Performance Toolbox (PTX) for AIX
      1. Performance Toolbox concepts
      2. Graphical monitoring and analysis issues
      3. Manager (1/2)
      4. Manager (2/2)
      5. Agent
      6. Monitoring an SMP with the performance toolbox
  10. Appendix A: Special notices
  11. Appendix B: Related publications
    1. IBM Redbooks publications
    2. IBM Redbooks collections
    3. Other resources
    4. Referenced Web sites
  12. How to get IBM Redbooks
    1. IBM Redbooks fax order form
  13. Abbreviations and acronyms (1/2)
  14. Abbreviations and acronyms (2/2)
  15. Index (1/4)
  16. Index (2/4)
  17. Index (3/4)
  18. Index (4/4)
  19. IBM Redbooks review (1/2)
  20. IBM Redbooks review (2/2)

Product information

  • Title: Understanding IBM eServer pSeries Performance and Sizing
  • Author(s): Diana Gfroerer, Nigel Trickett, Tatsuhiko Nakagawa, Ravi Mani
  • Release date: February 2001
  • Publisher(s): IBM Redbooks
  • ISBN: 9780738419152