IBM eServer pSeries Cluster Systems Handbook

IBM eServer pSeries Cluster Systems Handbook

by Dino Quintero, Graeme Cassie, Mukesh Kumar Gupta, Edward EuiJoo Lee, See Poo Soh
Released October 2003
Publisher(s): IBM Redbooks
ISBN: None

Read it now on the O’Reilly learning platform with a 10-day free trial.

O’Reilly members get unlimited access to books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.

Start your free trial

Book description

The IBM eServer Cluster 1600 server, which was introduced to meet the rigorous demands of mission-critical enterprise applications, continues to offer outstanding performance, scalability, reliability, availability, serviceability, and management capabilities. In this IBM Redbooks publication, we highlight the benefits of using a Cluster 1600, and describe which hardware components can be managed by either Parallel System Support Programs (PSSP) Version 3, Release 5, or Cluster Systems Management (CSM) Version 1, Release 3, Modification 2.
This publication contains the following information on the Cluster 1600:
- Cluster 1600 hardware components
- Networking components and considerations
- Cluster 1600 software components
- Scalability of the Cluster 1600
- Solutions and offerings scenarios
The Cluster 1600 helps to reduce the complexities and costs of system management, thus lowering the total cost of ownership and allowing simplification of application service level management. It also provides the infrastructure that supports availability, data sharing, and response time. This book will be useful for IT professionals seeking to implement Cluster 1600 mission-critical solutions to address business intelligence applications, server consolidation, and collaborative computing.

Table of contents

  1. Notices
    1. Trademarks
  2. Preface
    1. Become a published author
    2. Comments welcome
  3. Chapter 1: Introduction
    1. Overview of Cluster 1600
    2. Choosing PSSP or CSM
      1. Cluster management with PSSP
      2. Cluster management with CSM (1/2)
      3. Cluster management with CSM (2/2)
  4. Chapter 2: Cluster 1600 hardware
    1. Overview
    2. Cluster 1600 hardware components
      1. Nodes
      2. Frames
      3. Switches
      4. PSSP control workstations
      5. CSM for AIX management server
      6. Hardware Management Console (HMC)
    3. CSM and PSSP hardware support
      1. CSM-managed node requirements
    4. PSSP control workstation
      1. Control workstation requirements
      2. Supported control workstations
      3. High Availability Control Workstation
      4. HACWS limitations
    5. CSM management server
      1. Memory and disk space
      2. Network requirements
      3. Asynchronous card requirements
      4. Using a Logical Partition (LPAR) as a CSM management server
    6. Cluster 1600 server concepts
      1. pSeries architecture (1/2)
      2. pSeries architecture (2/2)
      3. Cluster 1600 and the HMC (1/2)
      4. Cluster 1600 and the HMC (2/2)
      5. Firmware
      6. Electronic Service Agent
      7. Planning for Cluster 1600 servers
    7. Hardware supported and currently marketed
    8. pSeries servers
      1. pSeries 615 server (7029-6C3 and 6E3 deskside)
      2. pSeries 630 server (7028-6C4 and 6E4 deskside) (1/2)
      3. pSeries 630 server (7028-6C4 and 6E4 deskside) (2/2)
      4. pSeries 650 server (7038-6M2) (1/2)
      5. pSeries 650 server (7038-6M2) (2/2)
      6. pSeries 655 server (7039-651) (1/2)
      7. pSeries 655 server (7039-651) (2/2)
      8. 670 Server (7040-671) (1/2)
      9. 670 Server (7040-671) (2/2)
      10. 690 server (7040-681) (1/2)
      11. 690 server (7040-681) (2/2)
      12. xSeries servers
    9. Switches
      1. 9076 model 555
      2. 9076 model 556
      3. 9076 model 557
      4. 9076 model 558
      5. 7045-SW4 pSeries HPS (High Performance Switch)
    10. Switch adapters
      1. Switch adapter placement restrictions
      2. pSeries HPS switch network interface cards (SNI)
    11. Legacy hardware supported but no longer marketed
      1. pSeries 660 Model 6M1 (7026-6M1)
  5. Chapter 3: Network configuration
    1. SP LAN Ethernet
      1. Supported Ethernet adapters and their placement
      2. Ethernet network topology
      3. IP label convention
    2. Switch network
      1. Benefits of a Switch network
      2. SP Switch2
      3. Switch IP network and addressing
    3. Other networks
    4. Network considerations
      1. The RS-232 connection
      2. System topology considerations
      3. Boot/install server requirements
      4. The SP Ethernet administrative LAN
      5. Additional LANs - considerations
      6. IP over the switch - considerations
      7. Subnetting - considerations
      8. HMC trusted network - considerations
      9. Network router node considerations
      10. Clustered server configuration considerations
      11. SP-attached server considerations
    5. Sample scenarios of Cluster 1600 managed by PSSP
      1. CWS with two HMCs and four pSeries
      2. One CWS, one HMC and one pSeries
      3. CWS, two HMCs, two 9076s, with one pSeries and SP Switch2
      4. CWS, HMC, 9076 frame and pSeries with SP Switch
    6. Networking for Cluster Systems Management (CSM)
      1. CSM hardware control
      2. Hardware and network requirements
      3. Virtual LANs (VLANs)
      4. Conceptual diagram for pSeries cluster
      5. pSeries HPS switch network overview
      6. Switch Network Manager (SNM)
      7. Considerations for Cluster 1600 managed by CSM network
      8. Examples
      9. Redundant HMC Layout for pseries HPS in Cluster 1600
      10. Redundant layout for pSeries in the Cluster 1600
      11. Conceptual Cluster 1600 without a pSeries HPS
      12. Management Server with two HMCs and four pSeries
  6. Chapter 4: Software support
    1. Software components of the Cluster 1600
    2. Parallel System Support Programs (PSSP)
      1. Administration and operation
      2. Reliable Scalable Cluster Technology (RSCT)
      3. IBM Virtual Shared Disk (VSD)
      4. Security
      5. Communication subsystem
      6. Network Time Protocol (NTP)
      7. System availability
      8. Other PSSP services
      9. New in PSSP 3.5
      10. Software requirements
      11. Software compatibility matrix
      12. Documentation references - PSSP
    3. Cluster Systems Management (CSM)
      1. Administration and operation (1/2)
      2. Administration and operation (2/2)
      3. Reliable Scalable Cluster Technology (RSCT)
      4. New in CSM 1.3.2 for AIX
      5. Supported platform
      6. PSSP-to-CSM transition
      7. Documentation references - CSM
    4. General Parallel File System (GPFS)
      1. Architecture
      2. Administration and operation
      3. Higher performance/scalability
      4. Recoverability
      5. Migration
      6. New in GPFS 2.1 for AIX 5L
      7. Software requirements
      8. Documentation references
    5. LoadLeveler
      1. Administration and operations
      2. Capabilities
      3. New in LoadLeveler 3.2
      4. New in LoadLeveler 3.1
      5. Software requirement
      6. LoadLeveler configuration suggestions
      7. Documentation references - LoadLeveler
    6. Scientific subroutine libraries
      1. Engineering and Scientific Subroutines Library (ESSL) family of products
      2. Operations
      3. New in ESSL 4.1
      4. New in Parallel ESSL 3.1
      5. Software requirements
      6. Documentation references - ESSL and PESSL
      7. Mathematical Acceleration Subsystem (MASS)
    7. Parallel Environment (PE)
      1. Parallel Programming support
      2. Operation
      3. New in PE 4.1
      4. Software requirements
      5. Documentation references - Parallel Environment (PE)
    8. IBM High Availability Cluster Multi-Processing for AIX (HACMP)
      1. HACMP operations
      2. Administration and operation
      3. New in HACMP 5.1
      4. Software requirements
      5. Documentation references - HACMP
    9. Performance Toolbox (PTX) and Performance AIDE (PAIDE)
      1. Administration and operation
      2. Platform requirements
    10. Software ordering and configuration
  7. Chapter 5: Solutions and offerings “best practices”
    1. High Performance Computing (HPC) environment
      1. A hypothetical solution
      2. Solution architecture
      3. Solution discussion
      4. Cluster management considerations
    2. Transition from SP nodes to LPARs
      1. System migration by utilizing alternate disk migration
    3. Virtual Serial port implications with LPARS
      1. Console device
      2. Serial port implication in an LPAR/SP environment
      3. Virtual terminal window
    4. HMC considerations
      1. Redundant HMC
    5. Web-based System Manager client solutions
      1. Web-based System Manager functionality through the firewall
  8. Appendix A: Performance
    1. Switch performance
    2. Adapter performance
    3. Node I/O slot performance
    4. Application performance
    5. MPI/user space
    6. TCP/IP
  9. Related publications
    1. IBM Redbooks
    2. Other publications
    3. Online resources
    4. How to get IBM Redbooks
    5. Help from IBM
  10. Abbreviations and acronyms
  11. Index (1/2)
  12. Index (2/2)
  13. Back cover

Product information

  • Title: IBM eServer pSeries Cluster Systems Handbook
  • Author(s): Dino Quintero, Graeme Cassie, Mukesh Kumar Gupta, Edward EuiJoo Lee, See Poo Soh
  • Release date: October 2003
  • Publisher(s): IBM Redbooks
  • ISBN: None